Cvoice Vol 1-3

Table of Contents "
Volume 1
Course Introduction
1
Overview 1
Learner Skills and Knowledge 3
Course Goal and Objectives 4
Course Flow 5
Additional References 5
Cisco Glossary of Terms 6
Your Training Curriculum
1-1
Introduction to Voice Gateways
mm CiscoUnified Communications Gateways
Gateway Functionality
VoIPSignalingProtocols
*" Gateway Deployment Example
Gateways in Cisco Unified Communications Deployment Models i-
Single-Site Deployment ]~^j
Multisite WAN with Centralized Call Processing -
* Multisite WAN with Distributed Call Processing ]'&
*"* Clustering over the IP WAN
Gateway Hardware Platforms
Gateway Operational Modes
J" Voice Gateway Call Legs
" Voice-Switching Gateway J"^
VoIP Gateway ]"
Cisco Unified Border Element j"j'
t Summary " _
* References
Examining Gateway Call Routing and Call Legs 1^2
Objectives^ ^ ^ ^ ]^
1-41
1-42
1-43
1-44
1-45
1-46
1-47
1-49
1-52
1-53
Gateway Call-Routing Components
Most Prevalent Dial-Peer Types
Dial Peers
VoIP Dial Peers
VoIP Dial Peer Examples
End-to-End Call Routing
Call Routing
Call Legs
Configuring POTS Dial Peers
Dial Peer Matching
String-Matching Characters
Number-Matching Characters ]*^
Matching Inbound Dial Peers !"^
Matching Outbound Dial Peers y
Default Dial Peer
Direct Inward Dialing ]jj*j
Two-Stage Dialing
One-Stage Dialing 1"71
1"1
Overview ^
Module Objectives
Understanding Cisco Unified Communications Networks and theRole of Gateways 1-3
Objectives ,'_^
Cisco Unified Communications
Cisco Unified Communications Overview
Cisco Unified Communications Architecture
Cisco Unified Communications Business Benefits Vjj
1-11
1-13
1-18
1-6
1-7
1-26
1-28
1-33
1-34
Summary
Configuring Gateway Voice Ports
Objectives
1-75
1-77
Voice Ports Overview 4~-,n
Voice Trunk Example
Installing Voice Ports
Analog Voice Ports
Analog Signaling Overview ,' gf
Analog Signaling
Address SignalingDTMF
Call Progress Tones
Configuring Analog Voice Ports ^gg
Configuring FXS Voice Ports 19q
Configuring FXO Voice Ports 1"91
Configuring FXS-DID Voice Ports 1.g2
Configuring E&M Voice Ports 1_93
Configuring CAMA Voice Ports 1g4
Digital Voice Ports 1g5
Digital Circuit Types 1 qfi
T1 CAS Overview
T1 CAS SF Format
T1 CAS ESF Format
E1 CAS Overview
1-78
1-81
1-82
1-83
1-85
1-87
1-E
1-98
1-99
1-100
E1 CAS Multiframe Format 1-102
Understanding ISDN 1 103
ISDN BRI and PRI Interfaces ^-,05
ISDN Architecture -j ^07
Non-Facility Associated Signaling 1_108
Configuring Digital Voice Ports 1-109
Configuring Digital Ports 1-110
Configuring ISDN 1-116
ISDN BRI Configuration -j 1ig
ISDNE1 PRI Configuration ^g
Fine-Tuning Analog and Digital Voice Ports 1_121
Fine-Tuning Analog Voice Ports 1-123
Echo Cancellation 1-126
Talker Echo 1-127
Listener Echo 1-128
Echo Cancellation 1-129
Echo Canceller Parameters 1_130
Configuring Echo Cancellation 1-132
Verifying Analog and Digital Voice Ports 1^33
showvoice port summary Command 1-134
Verifying Analog Voice Ports ^-^35
Verifying Voice Ports 1-136
Summary 1-142
Understanding DSP Functionality. Codecs, and Codec Complexity 1-143
Objectives 1-143
Voice Codecs -\-\AA
Voice Codec Packet Rates and Payload Sizes 1-148
Evaluating Quality ofCodecs 1_14g
Mean Opinion Score 1-149
Perceptual Evaluation of Speech Quality 1-149
Perceptual Evaluation of Audio Quality 1-150
CodecQuality 1-151
Evaluating Overhead 1-152
Per-Call Bandwidth Using CommonCodecs 1-156
Digital Signal Processors 1-157
DSP Modules -!_161
Implementing Cisco Voice Communications and QoS (CVOICE) v8.0 2010 Cisco Systems.
1 1fi9
DSPModule Comparison "
Codec Complexity 1 1fiI-
Packet Voice DSP Module Conferencing ^^
DSP Calculator ^_^7
Configuring DSPs . 1fifi
Configuring DSP Services for Voice Termination ' "
CodecComplexity Configuration "
Configuring DSP Resources for Transcoding, Conferencing, and MTP i-1/u
Transcoding and Conferencing Example ^73
Verifying DSPs 1175
Summary 1-175
References 1-177
Module Summary 1-177
References ^^9
Module Self-Check ".
Module Self-Check Answer Key
VoIP Call Legs
2-1
2-1
Overview 2 1
Module Objectives
Examining VoIP Call Legsand VoIP Media Transmission . ZA
" 2-3
Objectives 2_d
VoIP Overview - _
Major Stages of Voice Processing in VoIP *-?
VoIPComponents '
Converting Voice to VoIP
Sampling
Quantization
>2010 Cisco Systems, Inc. Implementing Cisco Voice Communications and QoS (CVOICE) v8.0
2-8
2-9
2-12
Encoding 13
Compression
VoIP Packetization
Packetization Rate
Codec Operations
Packetization and Compression-G.729 Example
VoIP Media Transmission ^"Jn
Real-Time Transport Protocol
Real-Time Transport Control Protocol
Compressed RTP
Secure RTP
Secure RTP Packet Format j~jz
VoIP Media Considerations ^"^
Voice Activity Detection
Bandwidth Savings
Voice Port Settings for VAD
Summary
Explaining H.323 Signaling Protocol
Objectives
H.323 Architecture
H.323 Advantages
H.323 Network Components
H.323 Gateways
H.323 Gatekeepers
H.323 Multipoint Control Units
H.323 Multipoint Conferences
H.323 Regional Requirements Example
2-14
2-15
2-16
2-17
2-19
2-20
2-21
2-22
2-27
2-28
2-29
2-30
2-31
2-31
2-32
2-33
2-35
2-36
2-37
2-38
2-39
2-40
H.323 Call Flows 2-41
H.323 Slow Start Call Setup
H.323 Slow Start Call Teardown f
H.225 RAS Call Setup 2-44
H.225 RAS Call Teardown
Codecs in H.323
Negotiation in Slow Start Call Setup
H.323 Fast Connect
H.323 Early Media
Configuring H.323 Gateways
H.323 Gateway Configuration Example
Customizing H.323 Gateways
H.323 Session Transport
Idle Connection and H.323 Source IP Address
H.225 Timers
H.323Gateway Tuning Example
Verifying H.323Gateways
Summary
Explaining SIP Signaling Protocol
Objectives
SIP Architecture
Signaling and Deployment
SIP Architecture Components
SIP Servers
SIP Architecture Examples
SIP Call Flows
SIPCall Setup Using Proxy Server
SIP Call Setup Using Redirect Server
SIP Addressing
Address Registration
Address Resolution
Codecs in SIP
SDP Examples
Delayed Offer
Early Offer
Early Media
Configuring Basic SIP
User Agent Configuration
Dial Peer Configuration
Basic SIP Configuration Example
Configuring SIP ISDNSupport
Calling Name Display
Calling Name Display Commands
Calling Name Display Configuration
Slocking and Substituting Caller ID
Blocking and Substituting Caller ID Commands
Blocking and Substituting CallerID Configuration
Configuring SIP SRTP Support
SIPS Global and Dial Peer Commands
SRTP Global and Dial Peer Commands
SIPS and SRTP Configuration Example
Customizing SIP Gateways
SIP Transport
SIP Source IP Address and UA Timers
SIP UA Timers
SIP Early Media
Gateway-to-Gateway Configuration Example
UA Example
Verifying SIP Gateways
SIP-UA General Venfication
SIP-UA Registration Status
SIP-UA Call Information
SIP Debugging Overview
Examining the INVITE Message
Implementing Cisco Voice Communications and QoS (CVOICE) v8 0
2-46
2-47
2-48
2-49
2-50
2-51
2-52
2-53
2-54
2-55
2-56
2-57
2-58
2-59
2-61
2-61
2-62
2-64
2-65
2-66
2-67
2-68
2-70
2-71
2-72
2-73
2-74
2-75
2-76
2-77
2-78
2-79
2-81
2-82
2-83
2-84
2-85
2-86
2-87
2-88
2-89
2-90
2-91
2-92
2-93
2-94
2-95
2-96
2-97
2-98
2-99
2-100
2-101
2-102
2-103
2-104
2-105
2-106
2-107
2-109
2010 Cisco Systems, Inc.
_ 2-110
Examining the 200OK Message 2-111
Examining the BYE Message 2-112
Summary 2-113
m Explaining MGCP Signaling Protocol ^77^
Objectives 2-114
MGCP Architecture 2-115
MGCP Key Features 2-117
MGCP Components 2-119
MGCP Gateways 2-120
MGCP Endpoints 2-121
MGCP Package Types 2-122
MGCP Call Flows 2-125
Residential Gateway toResidential Gateway 2127
Trunking Gateway toTrunking Gateway 2-128
MGCP Special Considerations _ Qmni0\ 2-129
m Codec Negotiation (Residential Gateway-to-ResidenHal Gateway Example) 2129
Digit Collection 2-131
Configunng MGCP Gateways 2A32
MGCP Commands 2-133
mm Customizing MGCP Gateways 2134
Package Configuration 2-135
SelectedPackageTypes 2-136
Residential Gateway Example 2-137
w Trunking Gateway Example 2-138
Verifying MGCP Gateways 2_139
show mgcp Command 2-140
show ccm-manager Command 2-141
mm show mgcp endpoint Command 2-142
show mgcp statistics Command 2-143
Summary
* Describing Requirements for VoIP Call Legs *=1*|
Objectives 2_146
Audio Clarity 2-148
*. Delay 2_149
Delay Types 2_150
^ Acceptable Delay (G.114) 2152
Jitter 2^ 53
Packet Loss 2 154
r Bandwidth Requirements 2155
""* QoS Requirements 2-157
QoS Objectives 2 158
Transporting Modulated Data over IP Networks
J_ Differences from Fax Transmission in the PSTN 5g
, Fax Services over IP Networks
Understanding FAX/Modem Pass-Through, Relay, and Store and Forward 2-ibU
Pass-Through Topology 2]163
f Pass-Through 2_165
" Relay Topology 2166
Relay ,. 2-168
Store-and-Forward Fax fiq
S: Gateway Signaling Protocols, and Fax and Modem Pass-Through and Relay 2-iby
1mm Cisco Fax Relay 2 173
T.38 Fax Relay 2 179
DTMF Support 2 180
6 DTMF Support 2-184
^" Summary
2010 Cisco Systems, Inc. Implementing Cisco Voice Communications and QoS (CVOICE) v8.0
Configuring Vp|p Call Lpg<t
Objectives
2-185
Configuration Components ofVoIP Dial Peer
VoIP Dial Peer Characteristics 2~186
Configuring DTMF Relay 2"187
DTMF Relay Configuration Example 2"188
Configuring FAX/Modem Support 2~189
T.38 Fax Relay Configuration 2~190
Fax Relay Speed Configuration 2~191
Fax Relay SG3 Support Configuration 2"]92
Fax Support Configuration Example i']T*
Configuring Modem Support ^~194
Modem Pass-Through and Modem Relay Interaction I'lll
Modem Support Configuration Example i~\zz
Configuring Codecs 2-199
Codec-Related Dial Peer Configuration 2"^?
Codec Configuration Example *~zil
Limiting Concurrent Calls 2"202
Summary 2-203
Module Self-Check 2"205
Module Self-Check Answer Key 2"2^
Implemenling Cisco Vorce Communications and QoS (CVOICE! v8.0 2010 Cisco Systems Inc
CVOICE
Course Introduction
Overview
Implementing Cisco Ioice Communications and QoS (CVOICE) v8.0 teaches learners about
voice gateways, characteristics of VoIP call legs, dial plans and their implementation, bastc
implementation of IP phones in aCisco Unified Communications Manager Express
environment, and essential information about gatekeepers and Cisco Unified Border Element.
The course provides learners with voicc-relatcd quality of service (QoS) mechanisms, which
are required in Cisco Unified Communications networks.
Learner Skills and Knowledge
This subtopic lists the skills and knowledge that learners must possess to benefit fully from this
course. The subtopic also includes recommended Cisco learning offerings that learners should
first complete before taking the course.
Learner Skills and Knowledge
Working knowledge of fundamental terms and concepts of
computer networking, including LANs, WANs, and IP
switching and routing
Ability to configure and operate Cisco IOS routers in an IP
environment at the Cisco CCNA Routing and Switching level
Basic knowledge of traditional voice, converged voice, and
data networksat the Cisco CCNA Voice level
Learner Skills and Knowledge (Cont)
Cisco learning offerings:
Introducing Cisco Voice and Unified Communications
Administration (ICOMM)
Implementing Cisco Voice Communications andQoS (CVOICE] v8 0
2010Cisco Systems, Inc
Course Goal and Objectives
This topic describes the course goal and objectives.
"To provide learners with the necessary knowledge to
implement and operate gateways, gatekeepers, Cisco
Unified Border Element, Cisco Unified Communications
Manager Express, and QoS in a voice network
architecture''
>rrp:wm-rg Qscc Vo.ce Communicate and QoS .CVOICE) 8.0
Upon completing this course, you will be able to meet these objectives:
. Explain what avoice gateway is, how it works, and describe its usage, components, and
features
Describe the characteristics and configuration elements of VoIP call legs
. Describe how to implement IP phones using Cisco Unified Communications Manager
Express
. Describe the components of adial plan, and explain how to implement adial plan on a
Cisco Unified voice gateway
Explain what gatekeepers and Cisco Unified Border Elements arc, how they work, and
what features they support
. Describe why QoS is needed, what functions it performs, and how it can be implemented in
a Cisco Unified Communications network
i 2010 Cisco Systems, Inc.
Course Introduction
Course Flow
This topic presents the suggested ilow ofthe c
course materials.
tmethMh- mmended structure for this course. This structure allows enough
" I " I0 ^T, "K Kl,rSe 'n,0nmti0n md fr >10 rk *B" the lab^
dem tics. Ihe exact timing of the subject materials and labs depends on the pace of vour
^p^cinc ciuss.
2010 Cisco Systems, Inc
Additional References
This topic presents the Cisco icons and symbols that are used in this course, as well as
information on where to find additional technical references.
Cisco Icons and Symbols
Cisco Unified
Presence
Cisco Unrty
Connection
Csco Unified
Messaging
Gateway
CecoASA
Ad*(^Security
Appliance
Cisco Unified
commrni cations
Manager
Cisco Unified
Border Element
Cisco Unified
Personal
Communicattr
Cisco unified
SRST Router
SAF- Erabled
Router
Network
Cloud
Gatekeeper
Switch Router
Cisco Unified
Communi cations
Manager Express
Cisco Unified
Communications
Manager Expresswith
Cisco UnrtyExpress
Cisco Glossary of Terms
For additional information on Cisco terminology, refer to the Cisco Internetworking Terms and
Acronyms glossary at /ttai
hUp:ôcvuki.c1sco.com/,,ki/Calegor>:tn{emetwt)rk,ng.Jem,sând_AcronymsjnA).
Course Introduction
Your Training Curriculum
This topic presents the training curriculum for this course.
You are encouraged to join the Cisco Career Certification Community, adiscussion forum that
r^;.to i Ul, ls ldi^ avalid nCareer Certification (such as Cisco CC
CCNA.CC DA" CCNP". CCDP\ CCIP\ CCVP\ or CCSP"). It provides agating place
or ,Sco cert-tied profes.onals to share uuestions. suggestions, and information about CisT
Caitcr Ccrtihcat.on programs and other certification-related topics. For more information visit
"Up: uuu.^Nco.tvm tin certifications.
)2010 Cisco Systems, Inc
Cisco Career Certifications: CCNP Voice
Expand your professional options and advance your career
Professional-level recognition in voice networking
Expert
Professional
Associate
Voice Nefwwking
Recommended TnitoWgThrough
Ctsco teaming Partners
Implementing Cisco Volet Commanicalkms
and QOS
imptomonting Cisco UnmH Commuricatons
Manager, Pm 1
trnpiamomm Cisco UMffiW CotmartcaSom
rmi*SfW(#nj Ciscoumt
Commurtcafcm
integrating oaoo Unified Cenwiuntoafiono
Applications
hiIp.//www.cisco.com/go/certitications
Course Introduction
Implementing Cisco Voice Communications and QoS (CVOICE) v8 0 2010 Cisco Systems. Inc.
Module 11
Overview
Cisco Unified Communications gateways play an important role in the Cisco Unified
Communications environment. Their primary function is to convert voice formats, s.gnals, and
transmission methods as voice information travels over various network types.
This module describes the various types of voice gateways and how to deploy them in different
Cisco Unified Communications environments. Furthermore, itexplains the call routing process,
the direct inward dialing (DID) feature, the various types ofvoice ports and their
characteristics, coder-decoders (codecs), digital signal processors (DSPs), and their
implementation.
Module Objectives
Upon completing this module, you will be able to explain what avoice gateway is. how it
works, and describe its usage, components, and features. This ability includes being able to
meet these objectives:
Describe the characteristics and historical evolution ofunified communications networks,
the three operational modes of gateways, their functions, and the related call leg types
Fxplain how gateways route calls and which configuration elements relate to incoming and
outgoing call legs
Describe how to connect agateway to traditional voice circuits using analog and digital
interfaces
Define DSPs and codecs, and explain different codec complexities and their usage
1-2 Implementing Cisco Voice Communications and QoS (CVOICE) v8 0
2010 Cisco Systems. Inc.
Lesson 1
Understanding Cisco Unified
Communications Networks
and the Role of Gateways
Overview
This lesson describes the operational modes of a voice gateway and how it fits intheCisco
Unified Communications architecture. It explainsthe voicegateway functions ineach Cisco
Unified Communications deployment model andthecall legsthat areassociated with each
operational mode.
Objectives
Upon completing this lesson, you will beable toexplain what a voice gateway is, how it works,
and describe its usage, components, and features.
This abilityincludes beingable to meet these objectives:
Describe the architecture and components of CiscoUnified Communications architecture
Describe the function of voicegateways and their major roles in Cisco Unified
Communications networks
Identify theroleof thegateways infour supported Cisco Unified Communications
deployment models
Briefly describethe differentCiscovoice gateway platforms
Identity the call legsthat are createdby a voice gateway in eachoperational mode
1-4
1his topicdescribes the Cisco Unified Communications architecture and its evolution from
traditional telephony.
Traditional Telephony Networl
i ns
[Wees
*'-* Siwinj'" "Swicn"
r>/
San Jose
\V
PSTN
n
1 s
Boston
The figure illustrates the typical components of a traditional telephony network:
Telephones: Analog telephones arc the most common type of phone in a traditional
telephony network. Analogphones directlyconnect to the publicswitched telephone
network (PSTN).
Central office (CO) switch: These switchesterminate the local loopand manage
signaling, digil collection, call routing, call setup, and call teardown.
PBX: A PBX is a privately ownedswitch that is located on the customerpremises. A PBX
is a smaller, privately owned version of the COswitches that telephone companies (teleos)
use. Many businesses still have a PBXtelephone system. Largeoffices with more than 50
telephones or handsets still use a PBX to connect users, both in-house and to the PSTN.
Trunk: Trunks provide the path between two switches and can be of different types:
CO trunk: A CO trunk is a direct connection between a local CO and a PBX. which
can be analog or digital.
Tie trunk: A tie trunk is a dedicated circuit that connects PBXs to each other.
Interoffice trunk: An interoffice trunk is typically a digital circuit that connects the
COs of two local teleos.
Traditional telephony differs in many aspects from modem unified communications. One
important difference is the closed nature of traditional telephony. Integration with modem
software applications, databases, and a rapidly evolving computing environment is difficult.
Traditional telephony uses circuit-switching technology to establish a voice channel end-to-end.
Ihis approach does not allow sharing of the network infrastructure for emerging applications
and services.
Implementing Cisco Voice Communications and QoS (CVOICEI u8 0 ) 2010 Cisco Systems. Inc
Atraditional telephony environment addresses these areas:
Signaling- Signaling is the ability to generate and exchange the control information that
will be used to establish, monitor, and release connections between two endpoints. Voice
signaling requires the ability to provide supervisory, address, and alerting functionality
between nodes The PSTN network uses Signaling System 7(SS7) to transport control
messages. SS7 uses out-of-band signaling, which, in this case, is the exchange or call
control information ina separate dedicated channel.
Database services: Database services include access to billing information, caller name
(CNAM) delivery, toll-free database services, and calling-card services. An example is
providing acall notification service that places outbound calls with prerecorded messages
at specific times to notify users of such events as school closures, wakeup calls, or
appointments,
Bearer control: Bearer control defines the bearer channels that carry voice calls. Proper
supervision of these channels requires that the appropriate call connect and call disconnect
signaling is passed between end devices. Correct signaling ensures that the channel is
allocated to the current voice call and that the channel is properly deallocated when either
side terminates the call. Connect and disconnect messages are carried by SS7 mthe PSTN
network.
2Q10 Cisco Systems, Inc. Introduction to Voice Gateways 1-5
Cisco Unified Communications Overview
This subtopic provides an overview ofCisco Unified Communications.
* Integrated solution
Includes voice, video, data, and mobile applications
Builds on CiscoBorderless Networks as a secure network
architecture for all communications
MtUfr
IB*
J-rL-Muia M**> *.
"** .V.!
The Cisco Unified Communications system fully integrates communications by enabling data
voice. and video lo be transmitted over asingle network infrastructure using standards-based
IP. Ihe Cisco Unified Communications system incorporates and integrates the following
communications technologies:
IP communications is the technology that transmits voice and video communications over a
network using IPstandards. Cisco Unified Communications includes hardware and
software products, such as call-processing agents. IP phones (both wired and wireless),
voice-messaging systems, video devices, and many special applications.
Mobile applications enhance access to enterprise resources, increase productivity, and
increase the satisfaction of mobile users,
Customer care enables efficient and effective customer communications across aglobally-
capable network. This strategy allows organizations to draw from abroader range of
resources to sen-ice customers. They include access to alarge pool ofagents and multiple
channels ofcommunication, aswell ascustomer self-help tools.
1elepresence and conferencing enhance the virtual meeting environment with an integrated
set ofIP-based tools for voice, video, and web conferencing.
Messaging prov ides the functionality for sending and managing ofvoice and video
messages for users.
F.nterprise social software includes applications that enable communications with the
enterprise that are not strictly limited to business-orientedactivities.
Implementing Cisco Voice Communications and OoS(CVOICE) v80
Cisco Unified Communications Architecture
This subtopic describes the Cisco Unified Communications architecture.
Leveraging the framework provided by Cisco IPhardware and software products, the Cisco
Unified Communications system has thecapability to address current andemerging
communications needs inthe enterprise environment. TheCiscoUnified Communications
family of products isdesigned tooptimize feature functionality, reduce configuration and
maintenance requirements, and provide interoperability with a wide variety of other
applications. The Cisco Unified Communications system consists of these logical layers:
Infrastructure: Infrastructure consists of Cisco network components. It provides and
maintains a highlevel of availability, quality of service (QoS), andsecurity for the
network.
Services: Services areresponsible for providing thecore functionality of Cisco Unified
Communications, such as signaling and call routing.
Applications: Applications include a wide array of software thatoffers rich features to the
users.
Endpoints: Hndpoints include end-user hardware and software products that constitute
attachment pointsto the Cisco Unified Communications system.
) 2010 Cisco Systems. Inc.
Introduction to Voice Gateways 1-7
Cisco Unified Communications Business Benefits
This subtopic describes the Cisco Unified Communications business benefits.
Cisco Unified Communication;
Business Benefits
Cost savings
Flexibility
Advanced features
Advanced call routing
Unified messaging
Integrated information systems
Long-distance toll bypass
Voice security
Customer relationship
Telephony application services
Telepresence
Conferencing
The business advantages that influence the implementation of Cisco Unified Communications
have changed over time. Starting with simple media convergence, these advantages have
evolved to include call-switching intelligenceand the total user experience.
Originally, return on investment (ROI) calculations centered on toll-bypass and converged-
network sa\ ings. Although these savings are still relevant today, advances in voice
technologies alloworganizations and service providers to differentiate their product offerings
by providing these advanced features:
Cost savings: Traditional time-division multiplexing (fDM). which is used in the PSfN
environment, dedicates 64 kb/s of bandwidth per voice channel. This approach results in
unused bandwidth when there is no voice traffic. VoIP shares bandwidth across multiple
logical connections, which makes more efficient use of the bandwidth and therefore
reduces bandwidth requirements. A substantial amount of equipment is needed to combine
64-kb/s channels into high-speed links for transport across the network. Packet telephony
uses statistical analysis to multiplex voice traffic alongside data traffic. This consolidation
results in substantial savings on capital equipment and operations costs.
Flexibility: The sophisticated functionality of IP networks allows organizations to be
flexible in the types of applications and services that they provide to their customers and
users. Service providers can easily segment customers. This segmentation helps them to
providc different applications, custom services, and rates, depending on the traffic volume
needs and other customer-specific factors.
Implementing Cisco Voice Communications and QoS (CVOICE) v8 0 ) 2010 Cisco Systems, Inc.
Advanced features: Here are some examples ofthe advanced features provided by Cisco
Unified Communications:
Advanced call routing: When multiple paths exist toconnect acall toits
destination, some of these paths may bepreferred overothers based oncost,
distance, quality, partner handoffs, traffic load, orvarious other considerations,
Least-cost routing and timc-of-day routing aretwoexamples of advanced call
routing that can beimplemented lodetermine the best possible route for each call.
Cnified messaging: Unified messaging improves communications and productivity.
Itprovides asingle user interface for messages that have been delivered over various
media. For example, users can read their email, hear their voice mail, and view fax
messages by accessing a single inbox.
Integrated information systems: Organizations use Cisco Unified
Communications toaffect business process transformation. These processes include
centralized call control, geographically dispersed virtual contact centers, andaccess
to resources and self-help tools.
Long-distance toll bypass: Long-distance toll bypass isanattractive solution for
organizations thatplace a significant number of calls between sites that arccharged
traditional long-distance fees. Inthiscase, it may bemore cost-effective touse VoIP
toplacethose callsacross the IPnetwork. If the IPWAN becomes congested, calls
can overflowinto the PSTN, ensuring that there is no degradation in voice quality.
Voice and video security: There are mechanisms inthe IP network that ensure
secureIPconversations. Encryption of sensitive signalingheaderfields and message
bodies protects thepackets incaseof unauthorized packet interception.
Customer care: The abilityto providecustomersupport through multiple media,
such as telephone, chat, andemail, builds solidcustomer satisfaction andloyalty. A
pervasive IPnetwork allows organizations to provide contact centeragents with
consolidatedand up-to-date customer records along with the related customer
communication. Access to this informationallows quick problem solving, which, in
turn, builds strong customer relationships.
Telepresence and conferencing services: Theseservices savetime and resources
by providing a media-rich communications platform forusers ina distributed
enterprise environment.
2010 Cisco Systems, Inc. Introduction to Voice Gateways 1-9
Cisco Unified Communications Gateways
This topicdescribes the roles and functionality of gateways inthe CiscoUnified
Communications svstem.
Gateway Functionaiity
Unified communication gateways
connect voice-enabled
communication networks together.
Specifically, they can fulfill these
tasks:
Switch voice channels between
connected analog and digital
voice circuits
Convert voice formats between
traditional and VoIP networks
Interconnect two logically
separate VoIP networks
Unified communications gateways are connection points between different communications
networks. Depending on the deployment type, a gateway can perform one or several of these
functions:
Act as a voice switch that interconnects multiple traditional telephony circuits. The circuits
can be analog or digital. The gateway participates in signaling and may have to convert the
media channels. Gateways provide physical access for local analog and digital voice
devices such as telephones, fax machines, key sets, and PBXs.
Act as a PS I N-to-VolP gateway that provides translation between VoIP and non-VoIP
networks, such as the PSTN. In addition to the functionality of traditional voice switches,
the PS l'N-to-IP gateways enable voice and video communications between traditional
PSTN infrastructure and converged IP networks.
Act as a Cisco Unified Border [ lenient that interconnects two IP networks and allows
communications between endpoints distributed among them. The Cisco Unified Horder
Flemenls may implement filtering, address translation, and security-related functions.
Implementing Cisco Voice Communications and QoS (CVOICE) v8 0 >2010Cisco Systems. Inc.
Gateway Functionality
This subtopic describes the gateway functionality.
Gateway Functionality (Cont.)
Supports these VoIPsignaling protocols:
- H.323
- MGCP
- SIP
- SCCP
Works with redundant Cisco Unified Communications
Managers
Enables call survivability
Provides analog/digital interfaces to a PBX and the PSTN
Provides fax/modem services
Cisco Unified Communications gateways support these signaling protocols:
H.323: H.323 is a standard that specifies the components, protocols, and procedures that
provide multimedia communication servicesreal-time audio, video, and data
communicationsover packet networks, including IP networks. H.323 is part of a family
of ITU-T recommendations called H.32x that provides multimedia communication services
over a variety of networks. It is actually an umbrella of standards that define all aspects of
synchronized voice, video, and data transmission. It also defines end-to-end call signaling.
MGCP: Media Gateway Control Protocol (MGCP) is a method for PSTN gateway control
or thin device control. Specified in RFC 2705, MGCP defines a protocol that controls VoIP
gateways that arc connected to external call control devices, referred to as call agents.
MGCP provides the signaling capability for less-expensive edge devices, such as gateways,
that may not have a complete voice-signaling protocol such as H.323 implemented. For
example, any time an event such as off hook occurs at the voice port of a gateway, the
voice port reports that event to the call agent. The call agent then signals that device to
provide a service, such as dial tone signaling.
SIP: Session Initiation Protocol (SIP) is a detailed protocol that specifies the commands
and responses to set up and tear down calls. SIP also details features such as security,
proxy, and transport control protocol (TCP or User Datagram Protocol fUDP]) services.
SIP and its partner protocols. Session Announcement Protocol (SAP) and Session
Description Protocol (SDP), provide announcements and information about multicast
sessions to users on a network. SIP defines end-to-end call signaling between devices. SIP
is a text-based protocol that borrows many elements of HTTP, using the same transaction
request-and-response model and similar header and response codes. It also adopts a
modified form of the URL addressing scheme that is used within email, which is based on
Simple Mail Transfer Protocol (SMTP).
>2010 Cisco Systems, Inc. Introduction to Voice Gateways 1-11
SCCP: Skinny Client Control Protocol (SCCP) is a Cisco proprietary protocol used
between Cisco Unified Communications Manager and Cisco Unified IP phones. Cisco
Unified IP phones that use SCCP arc called Skinny clients. The client communicates with
Cisco Unified Communications Manager using connection-oriented (TCP/IP-based)
communication to establish a call with another end station.
Cisco Unified Communications gateways provide highly available communications platforms
using several methods, such as support for redundant Cisco Unified Communications Manager
systems and taking over some of the functionality of Cisco Unified Communications Manager
when it is not reachable due to network failure. The gateways provide a wide array of advanced
features, such as support for fax and modem communications.
1-12 Implementing Cisco VoiceCommunications and QoS (CVOICE) v8.0 2010 Cisco Systems. Inc.
VoIP Signaling Protocols
This subtopic describes the VoIP signaling protocols.
H.323: Peer-to-Peer Gateway Architecture
Each gateway maintains the dial plan.
Q.931
HV>1 is asuite of protocols defined by the ITU for multimedia conferences over LANs. The
II 3^3 protocol was designed by the ITU-T and initially approved in February 1996. It was
developed as aprotocol that provides IP networks with traditional telephony functionality.
Today. H.323 is the most widely deployed standards-based voice and videoconferencing
standard for packet-switched networks.
The protocols specified by H.323 include the following:
H225 call signaling is used to establish aconnection between two H.323 endpoints. This
connection is achieved by exchanging H.225 protocol messages on the call-signaling
channel. The call-signaling channel is opened between two H.323 endpoints or between an
endpoint andthe gatekeeper.
H225 Registration. Admission, and Status (RAS) is the protocol between endpoints
(terminals and gateways) and gatekeepers. The RAS is used to perform registration,
admission control, bandwidth changes, and status and disengage procedures between
endpoints and gatekeeper. An RAS channel is used to exchange RAS messages. 1his
signaling channel is opened between an endpoint and agatekeeper before the establishment
of any other channels.
11 245 control signaling is used to exchange end-to-end control messages governing the
operation of the H.323 endpoint. These control messages carry information related to the
following:
Capabilities exchange
Opening and closing oflogical channels used to carry media streams
Flow-control messages
General commands and indications
>2010 Cisco Systems, Inc.
1-13
1-14
third" nTTitfTenvironme;;ts-H323 is ^^* ^ &^^ m^P^. ^
third-part^ 11..23 chenls. especially video terminals. Connections are configured between
dev ices using staticdestination IPaddresses.
Implementing Cisco Voice Communications and QoS (CVOICE) v8 0 2010 Cisco Systems Inc
VoIP Signaling Protocols (Cont.)
MGCP: Client Server Architecture
Cisco Unified Communications Manager maintains the dial
plan.
Residential Gateway
E&u
MGCP
Trunking Gateway
PSTN
Q.921
Q.931
MGCP is a client-server call control protocol builtoncentralized control architecture. 1his
centralized control architecture hasthe advantage of centralized gateway administration and
prov ides for largely scalable IP telephony solutions. All the dial plan information resides on a
separate call agent. The call agent, which controls the ports on the gateway, performs call
control. The gateway does media translation between the PSTN and the VoIP networks for
external calls. In a Cisconetwork, CiscoUnified Communications Managersystems function
as the call agents.
MGCP isa plain-text protocol used by call control devices tomanage IPtelephony gateways.
MGCP was definedunder RFC2705 (MGCP Version 1.0),whichwas updatedby RFC3660
(Basic MGCP Packages), and superseded by RFC 3435 (MGCP Version 1.0). which was
updated by RFC 3661 (MGCP Return Code Usage).
With this protocol, Cisco Unified Communications Manager controls individual voice ports on
the gateway. MGCP allows complete control ofthe dial plan from Cisco Unified
Communications Manager. It gives Cisco Unified Communications Manager pcr-port control
of connections to thePSTN, legacy PBX, voice-mail systems, plain oldtelephone service
(POTS) phones, and soon. This control is implemented bya series of plaintext commands sent
over UDP port2427between CiscoUnified Communications Manager andthe gateway.
A PRI and BRI backhaul is an internal interface between the call agent (such as Cisco Unified
Communications Manager) and Cisco gateways. It isa separate channel forbackhauling
signaling information. APRI backhaul forwards PRI Layer 3(Q.931) signaling information via
a TCP connection.
An MGCP gateway is relatively easy to configure. Because the call agenthas all thecall-
routing intelligence, you do not need toconfigure the gateway with the complete dial plan that
it would otherwise need.
) 2010 Cisco Systems, Inc Introduction to Voice Gateways 1-15
SIP: Peer-to-Peer GatewayArchitecture
Each gateway maintains the dial plan.
IETF RFC. ASCII-text-based, WWW logic.
SIP isa protocol developed by the Internet Engineering Task Force (IETF) Multiparty
Multimedia Session Control (MMUSIC) workinggroupas an alternative lo 11.323. SIPfeatures
arccompliant with IETF RFC 2543. published inMarch 1999. RFC 3261, published inJune
2002. and RFC3665. published in December 2003. SIP is a common standard based on the
logic of the World Wide Web and very simple to implement. SIP iswidely used with gatewavs
and proxy servers w[thin service provider networks for internal and end-customer signaling.
SIPis a pecr-lo-peer protocol whereuser agents(UAs) initiate sessions, like H.323. But unlike
H.323. SIP uses ASCII text-based messages tocommunicate. Therefore, you can easily
implement and troubleshoot it. andanalyze the incoming signaling traffic content.
Because SIP is a peer-to-peer protocol. Cisco Iinified Communications Manager doesnot
control SIPdevices, and SIPdevices do not registerwithCiscoUnified Communications
Manager. As with H.323 gateways, only the IPaddress is required on Cisco Unified
Communication* Manager for the communication between Cisco Unified Communications
Manager and the SIP voice gateway.
Implementing Cisco Voice Communications and QoS (CVOICE) v8.0 2010 Cisco Systems. Inc.
j^^_I
VoIP Signaling Protocols (Cont.)
SCCP: Client Server Architecture
Cisco Unified Communications Manager servers maintain the
dial plan.
SCCP
SCCP
FXS
SCCP is aCisco proprietary protocol that is used for communications between Cisco Unified
Communications Manager'and terminal endpoints. SCCP is astimulus protocol, meaning that
any event (such as the phone is on hook or off hook, buttons have been pressed, and so on)
causes amessage to be sent to Cisco Unified Communications Manager. Spec.fic instructions
are then sent from Cisco Unified Communications Manager back to the device to tell it what to
do about the event. Therefore, each press on aphone button causes data traffic between Cisco
Unified Communications Manager and the terminal endpoint.
SCCP is vvidelv used with Cisco Unified IP phones. The major advantage of SCCP within
Cisco Unified Communications Manager networks is the broadest range of features that Cisco
Unified IPphones support viaSCCP.
SCCP is asimplified protocol used in VoIP networks. Cisco Unified IP phones that use SCCP
can coexist in an H.323 environment. When used with Cisco Unified Communications
Manager, the SCCP client can intemperate with H.323-compliant terminals.
1-17
Gateway Deployment Example
This subtopic prov ides an example of how gateways are deployed in an enterprise with mulfipk
locations
Gateway Deployment Example
Headquarters
IP WAN
PSTN-
H 323,'SIP f
Gateway
Branch
Gateway sare usually deployed as edge devices on anetwork. Because they typically interface
with both the PSTN and the company WAN. they must have the appropriate hardware and
utilize the appropriate protocol for that network. The figure represents ascenario in which two
different ty pes ol gateways are deployed in two locations for VoIP and PSTN interconnections,
fhe headquarters uses aCisco Unified Communications Manager environment with agateway
that connects the headquarters network to the PSTN and to the IP WAN and directs the calls '
through one ofthem. The gateway can be MGCP-controlled by Cisco Unified Communications
Manager or it can use one of the peer-lo-pecr signaling protocolsH.323 or SIP. With pecr-to-
peer signaling, the gateway will communicate not only with the Cisco Unified Communications
Manager but also with the branch gateway. The signaling communication with the branch
gateway occurs over the IP WAN. The headquarters gateway may use (he same protocol
(11.323 or SIP) for signaling exchange with the Cisco Unified Communications Manager and
the branch gateway, or it may use one protocol with Cisco Unified Communications Manager
(for example H.323) and the other protocol with the branch gateway (for example SIP). Ifthe
gateways in the headquarters and the branch use ditTerent protocols (one uses 11.323, the other
uses SIP), aCisco Unified Border Element is needed to translate the signaling information from
one protocol to another.
Ihe branch uses Cisco Unified Communications Manager Express on the voice gateway and
controls the Cisco Unified IP phones via SCCP and SIP. The gatewav can use SIP orH.323
when communicating with the headquarters. ACisco Unified Border Element may be needed
mease the branch and the headquarters use different signaling protocols.
18 Implementing Cisco Voice Communications andQoS (CVOICE) v8.0
2010 Cisco Systems. Inc
Gateways in Cisco Unified Communications
Deployment Models
This topic describes the voice gateway functions in the four common IP telephony deployment
models.
Deployment Models
Gateways support these IP telephony deployment models:
Single-site deployment
. Multisite WANwithcentralized call processing
Multisite WAN withdistributedcall processing
Clusteringovet the IPWAN
Applications
Cisco Unified Communications can be deployed in four models. Each deployment model
differs in die type oftraffic that is carried over the WAN, the location ofthe call-processinf
agent, and the'size of the deployment. Cisco Unified Communications supports these
deployment models:
Single-site
Multisite with centralized call processing
Multisite with distributed call processing
Clustering over the IP WAN
Single-Site Deployment
This subtopic explains the gateway functions in the single-site deployment model
Overview
Manager servers, applications,
and DSP resources at the same
physical location
IP WAN used for data traffic only
* PSTN used for all external calls
PSTN-
Cisco Unified
Communications
Manager Cluster
The single-site model for Cisco Unified Communications consists of acall-processing agent
cluster located at asingle site, with no telephony services provided over an IP WAN.
An enterprise would typically deploy the single-site model over aLAN or metropolitan-area
network (MAN), which carries the voice traffic within the site. Gateway trunks that connect
directly to the PSTN process all external calls. Ifan IP WAN exists between sites, itis used to
carry data traffic only: no telephony serv ices are provided over the WAN.
Single-site deployment offers aself-contained approach. There is no dependency for service in
the event ofan IP WAN failure orinsufficient bandwidth, and there is no loss ofcall-
processing senice orfunctionality. The main benefits ofthe single-site model are as follows:
Lase of deployment.
Acommon infrastructure for a converged solution.
Simplified dial plan. The dial plan describes how toforward calls that are based onthe
calling and called number. Because all external calls arc sent out over the PS'I'N trunk, the
dial planis easy to implement.
1-20 Implementing Cisco Voice Communications and QoS(CVOICE) 1/8 0
Single-Site Deployment (Cont.)
Gateway Functions
Cisco Unified Communications gateway:
Uses H.323 or MGCP for PSTN gateways
- Maintains the dial plan with H.323 or SIP
- Receives instructions fromthe MGCP call agent
Uses a single best-quality codec for all endpoints (G.711).
Provides DSP resources for conferencing and media
termination
Offers appropriate services:
- HSRP for gateway high availability
QoS mechanisms
- Security
In the single-site model, the voice gateway fulfills these functions:
Uses 11.323 or MGCP for the PSfN. MGCP simplifies the configuration, because the call
agent maintains the dial plan. The gateway must maintain the dial plan if H.323 or SIP is
used. SIP is typically deployed when connecting to an Internet telephony service provider
(ITSP).
Uses the G.711 codec for all endpoints. This practice eliminates the need to convert one
codec to another.
Offers enough DSP resources for the required media termination and conferencing.
Provides highly available, fault-tolerant network service based on a common infrastructure
philosophy. It implements the recommended QoS mechanisms and provides secure
platform communications.
>2010 Cisco Systems. Inc
Multisite WAN with Centralized Call Processing
This subtopic explains the gateway functions in the multisite centralized IPtelephony
deployment model.
1-22
Multisite WAN with Centralized Cal
Processing
Overview
* Cisco Unified Communications
Manager at central site;
applications centralized or
distributed
* IP WAN carries voice traffic
and call control signaling
Call Admission Control
(limit number of calls per site)
SRST for remote branches
AAR used if WAN bandwidth is
exceeded
The multisite WAN deployment model with centralized call processing consists of a single call-
processing agent cluster that provides services for many remote sites and uses the IP WAN to
transport Cisco Unified Communications traffic between the sites. The IP WAN also carries
call control signaling between the central site and the remote sites. The figure illustrates a
typical centralized call-processing deployment, with a Cisco Unified Communications Manager
cluster as the call-processing agent at the central site and an IP WAN to connect all the sites.
The remote sites rely on the centralized Cisco Unified Communications Manager cluster to
process their call processing. Applications such as voice mail, Cisco Unified Presence,
interactive voice response (IVR) systems, and so on, are typically also centralized to reduce the
overall costs of administration and maintenance.
The primary path for call control signaling and voice traffic is the IP WAN. To avoid
oversubscribing the WAN links with voice traffic and deteriorating the quality of established
calls. Call Admission Control (CAC) must be implemented.
In addition to IP WAN. Cisco gateways provide the remote sites with PSTN access. When the
IP WAN is down, or if all the available bandwidth on the IP WAN has been consumed, the
calls from users at the remote sites will be automatically redirected over the PSTN. The Cisco
Unified Sun. ivable Remote Site Telephony (Cisco Unified SRST) feature is available for both
SCCP and SIP phones. It provides call processing at the branch offices for Cisco Unified IP
phones if they lose their connection to the Cisco UnifiedCommunications Manager cluster or if
the WAN connection is down. Automated alternate routing (AAR) allows the central site
endpoints to communicate with the endpoint in the remote locations via the PSTN network.
mm
Multisite WAN with Centralized Call
Processing (Cont.)
Gateway Functions
AppliesQoS to minimize delay between
Cisco Unified Communications Manager and remote
locations to reduce voice cut-through delays
At the remote sites, uses SRST, Cisco
Unified Communications Manager Express in SRST mode,
SIP SRST. and MGCPgateway fallbackto ensure call-
processingsurvivability inthe event ofa WAN failure
Runs HSRP for redundancy and high availability
termination
In the multisite model with centralized call processing, the voice gateway fulfills these
functions:
Uses H.323 or MGCP between Cisco Unified Communications Manager and the voice
gateway.
Minimizes delay between Cisco Unified Communications Manager and remote locations to
reduce voicecut-through delays. Ciscorecommends 150ms maximum one-way.
Runs Hot Standby Router Protocol (HSRP) to provide voice gateway redundancy.
Provides enough digital signal processor (DSP) resources fortherequired media
termination (converting voice wavelengths to VoIP packets) andconferencing. The DSPs
for conferencing aretypically located onthecentral sitegateway. TheDSPs aredescribed
in a later lesson.
At the remote sites, uses these Cisco Unified SRST features to ensure call processing
survivability intheevent of a WAN failure (these features can beactivated simultaneously
on die same Cisco voice gateway):
for SCCP phones, uses Cisco Unified SRST on a CiscoIOS gateway or Cisco
Unified Communications ManagerKxpress runningin SRSTmode
For SIP phones, uses SIP SRST
For MGCP phones, uses MGCPgateway fallback
) 2010 Cisco Systems, Inc.
Multisite WAN with Distributed Call Processing
1-24
This subtopic explains the gateway functions in the multisite IP telephony deployment model
withdistributed call processing.
Multisite WAN with Distribute*
Processing
Overview
Cisco Unified
Communications
Manager and
applications
located at each
site
IP WAN carries
intercluster call
control signaling
* Scales to
hundreds of sites
* Transparent use
ofthePSTNifthe
IP WAN is
unavailable
I .CiscoUnified '
CommunicationsL
Manager Cluster
Cisco Unified
Communications
Manager Cluster
Gatekeeper
Themodel for a multisite WAN deployment with distributed call processing consists of
multiple independent sites. Each site has its own call-processing agent cluster connected to an
IP WAN that carries voice traffic between the distributed sites.
Thedistributed call-processing site may consist of any of the following:
Asingle sitewith itsowncall-processing agent, which canbeoneof the following:
Cisco UnifiedCommunications Manager
Cisco Unified Communications Manager Express
Other IP PBX
A centralized call-processing site with its associated remote sites
A legacy PBX with a VoIP gateway
An IP WANinterconnects all the distributed call-processingsites. Typically, the PSTNserves
as a backup connection between the sites in case that the IP WAN connection fails or does not
have any more available bandw idth. Asite connected onlythrough the PSTN is a standalone
site and is not covered by the distributed call-processing model.
Multisitedistributed call processing allows each site lo he completely self-contained. In the
event of an IP WAN failure or insufficient bandwidth, the site does not losecall-processing
service or functionality. Cisco Unified Communications Manager simply sends all calls
between the sites across the PSTN.
Implementing Cisco Voice Communications and QoS (CVOICE) v8 0 2010 Cisco Systems, Inc.
Multisite WAN with Distributed Call
Processing (Cont.)
Gateway Functions
Participates in H.323 gatekeeper-based CAC
Usesa single low-bandwidth WAN codec tosave bandwidth
and minimize transcoding
Provides DSP resources for transcoding, conferencing, and
media termination
Applies QoS for low latency in the IP WAN toensuretimely
VolPforwarding
Runs HSRPforredundancy and highavailability
Inthemultisite WAN with distributed call control, theCAC is implemented by thegatekeeper
that isconnected totheWAN, Thegatekeeper acts asthebandwidth broker and simplifies the
call routing across the IP WAN.
Inthe multisite WAN with distributed call processing, the Cisco Unified Communications
gateway fulfills these functions:
Uses H.323 or MGCP between Cisco Unified Communications Manager andthegateway.
Participates inthe gatekeeper-based CAC for intcrsite calls through the IP WAN.
Gatekeepers require that H.323 isused for intersite VoIP signaling.
Uses only one low-bandwidth codec for media transport over the IP WAN. This approach
reduces the need to convert one codec to another, and saves bandwidth.
Provides sufficient DSP resources tosupport therequired media termination, conferencing,
and transcoding. The DSPs for conferencing and transcoding are typically located onthe
central site gateway.
Ensures timelyVoIPforwarding throughthe IP WAN.
Runs HSRP to provide redundancy andhigh availability.
Clustering over the IPWAN
This subtopic explains the gateway functions in the clustering over the IP WAN deployment
model.
Clustering over the IP
Overview
Applications and Cisco Unified Communications Manager
systems of the same cluster distributed over the IP WAN
* IP WAN carries intracluster communication in addition to call
signaling and media
- CallAdmission Control (limit number of calls per site)
AAR used if WAN bandwidth is exceeded
PSTN
SIP or SCCP
SIP or SCCP
Cisco supports Cisco Unified Communications Manager clusters over a WAN. In the clustering
over the IP WAN model, a single CiscoUnified Communications Managerclusterand its
subscriber servers are split across multiple sites connected via a QoS-enabled WAN. The IP
WANcarries intracluster control traffic in addition lo call signaling and voice traffic.
Aspecial requirement of this model is low latency through the IPWAN, The maximum one
way delay between any Cisco Unified Communications Manager servers for all priority Intra
clusterCommunications Signaling (ICCS) traffic should notexceed XO-ms round-trip time
(R'I'T). Delay for other ICCS traffic should be kept reasonable to providetimelydatabase
access.
CAC protects the IP WAN fromoversubscription with too many calls. When the IP WAN is
down or the maximum supported number of calls is reached, the AAR redirects the calls from
one site to another v ia the PSTN.
Implementing Cisco Voice Communications and QoS (CVOICE) v8.0 2010 Cisco Systems, Inc
Clustering over the IP WAN (Cont.)
Gateway Functions
Applies QoS for low latency in the IP WAN:
- 80-ms maximum RTTfor ICCS traffic between anytwo
Cisco Unified Communications Manager serversinthe
cluster.
The ICCS traffic types are classified as either priority or
best-effort. Priority ICCS traffic is marked with IP
Precedence 3(DSCP 24orPHB CS3). Best-effort ICCS
traffic is marked with IP Precedence 0 (DSCP 0 or PHB
BE).
- Expedited forwarding ofVoIP packets.
termination.
Runs HSRPforredundancyand highavailability.
In the clustering over the IP WAN model, H.323 or MGCP can be used for signaling between
Cisco Unified Communications Managerand the gateway.
The IP WAN must be engineered to timely forward the delay-sensitive traffic types.
The gateways should provide sufficient DSP resources for conferencing and media termination.
and use HSRP tooffer a redundant and highly available solution.
Gateway Hardware Platforms
This topic describes the Cisco gateway hardware platforms.
Gateway Hardware Platform:
Modern Enterprise Models
;iSco2S0O Series Routers
Cisco 3900 Series Roulers
Ihe figure depicts some of the modern enterprise models that are usuallv used within enterprise
networks.
Cisco 2900Series Integrated Services Routers
Ihe Cisco 2900 Series Integrated Serv ices Routers comprise four models: Cisco 2901
Integrated Serv ices Router. Cisco 2911 Integrated Sen ices Router. Cisco 292 I Integrated
Serv ices Router, and Cisco 295 1Integrated Serv ices Router. These Integrated Services Routers
Generation 2platforms are future-enabled with multicore CPUs, support for high capacity
DSPs for future enhanced video capabilities, high-powered service modules with improved
availability, and Gigabit Ethernet switching with enhanced Power over Ethernet (Pofi).
Additionally, anew Cisco IOS Software Universal Image and Services Ready Engine module
enables you to decouple the deploy ment ofhardware and software. This decoupling provides a
flexible technology foundation that can quickly adapt to evolving network requirements.
All Cisco 2900 Series Integrated Services Routers offer embedded hardware encryption
acceleration, voice- and video-capable DSP slots, optional firewall, intrusion prevention, call
processing, voice mail, and application services. In addition, the platforms support the industrv -
widest range of wired and wireless connectivity options such as TI/EI. xDSL. and both copper
and fiber Gigabit Ethernet.
Cisco 3900 Series Integrated Services Routers
The Cisco 3900 Series Integrated Serv ices Routers comprise two models: Cisco 3925 and Cisco
3945 Integrated Sen ices Routers. In addition to providing the functionality ofthe Cisco 2900
Series Routers, the Cisco 3900 Series Routers ol'ler superior performance and flexibility for
network deploy merits from small business offices to large enterprise offices, while providing
industry-leading investment protection.
1-28 Implementing Cisco Voice Communications andQoS (CVOICE) uB.O
Gateway Hardware Platforms (Cont.)
Weil-Known Older Enterprise Models
Cisco 2600 Series Routers Cisco 3800 Series Routers
Thefigure shows theenterprise models of Cisco modular access routers thathave voice
gateway capabilities. These modelsare well-known and widelyused.
Cisco2800Series Integrated Services Routers comprise four models: Cisco2801 Integrated
Seniccs Router. Cisco2811 Integrated Services Router, Cisco2821 Integrated Services
Router, andCisco 2851 Integrated Services Router. Theseriesmaintains support for most of
the more than 90 modules that are available for the Cisco 1700 Series Modular Access Routers.
Cisco 2600 Series Multisenice Platforms, and Cisco 3700 Series Multiservice Access Routers.
The Cisco 2800 Series routers can deliver simultaneous, high-quality, wire-speed services up to
multiple Tl/El or xDSL connections. Therouters offerembedded encryption acceleration and.
on the motherboard, voice DSP slots, as well as the following:
Intrusion prevention system(IPS) and firewall functions
Optional integrated call processing and voice-mail support
High-density interfaces for a wide range of wired and wireless connectivity requirements
Sufficient performance and slot density for future networkexpansion requirements and
advanced applications
Cisco3800 Series Integrated Services Routers also feature embedded securityprocessing,
significant performance and memory enhancements, and newhigh-density interfaces. These
features deliver the performance, availability, and reliability that are required lo scale mission-
critical security. IP telephony, business video, network analysis, and web applications in the
most demanding enterprise environments. The Cisco 3800 Series Routers deliver multiple
concurrent seniccs at wire-speed T3/E3 rates.
2010 Cisco Systems. Inc. Introduction to Voice Gateways 1-29
The Integrated Sen ices routing architecture of the Cisco 3800 Scries Routers is based on the
architecture of the Cisco 3700 Series Routers. The routers aredesigned loembed and integrate
security and voice processing with advanced wired and wireless services for rapid deployment
ol newapplications, including application layer functions, intelligent networkservices, and
converged communications. TheCisco3800 Series Routers supportthe bandwidth
requirements for multiple East Ethernet interfaces perslot. TDM interconnections, and fully
integrated power distribution to modules supporting 802.3af PoE. It also supports the existing
portfolio of modular interfaces. This accommodates network expansion or changes in
technology as newsen ices and applications are deployed. Byintegrating the functions of
multiple separate dev ices into a single compact unit, the 3800 Series Router reduces the cost
and complexity of managing remote networks.
The Cisco 3800 Series models include the Cisco 3825 IntegratedServices Router and the Cisco
3845 Integrated Serv ices Router, available withthreeoptional configurations for ACpower.
AC power with integrated inline power support, and DCpower.
Implementing Cisco Voice Communications and QoS (CVOICE) v8.0 2010 Cisco Systems, Inc.
Gateway Hardware Platforms (Cont.j
Special Voice Gateways
Cisco AS5350XM
Seres Gateways
Cisco A554C0XM
Series Gateways
Cisco VG24B Gateway
Cisco 7200 Series Routers
To fit special needs within thecustomer unified messaging system, Ciscooffers standalone
voice gateways for specific purposes. Each of these voice gateways fulfills adifferent need,
such as theintegration of analog devices intotheunified messaging system, enhanced
performance, business-class functionality, adaptability, serviceability, and manageability.
Cisco ATA 186
TheCisco Analog Telephone Adaptor 186 (Cisco ATA 186) is a handset-to-Ethernet adapter
that allowstraditional telephone devicesto function as VoIPdevices. Customers can use IP
telephony applications byconnecting theiranalog devices toanalog telephone adapters,
The Cisco ATA I86supports twovoice ports, eachof which has an independent telephone
number anda single I0BASE-T Ethernet port. This adapter canmake useof existing Ethernet
LANs, in addition to broadband pipes suchas DSL, fixed wireless, and cable modem
deployments.
The CiscoATA180Seriesproductsare standards-based IPcommunications devicesthat
deliver VoIP terminations to businesses and residences.
Cisco VG248 Analog Phone Gateway
Cisco VG248 Analog Phone Gateway provides support for traditional analog devices while
taking advantage of thenewcapabilities that CiscoUnified Communications affords. TheCisco
VG248 AnalogPhone (iateway offers 48 fully featured analogports for use as extensions to the
Cisco Unified Communications Manager system in a compact 19-inchrack-mount chassis.
>2010 Cisco Systems, Inc. Introduction to Voice Gateways
Cisco AS5350XM Series Universal Gateway
The CiscoAS5350XM Gateway is theonly one-rack-unit gateway that provides data, voice,
andfax seniccs. as weli as Session Border Controller (SBC) functionality. TheSBC feature is
usedat prov ider interconnects and typically provides complete sessionstate, security, and
reporting sen ices. Ihe CiscoAS5350XM offershighreliability in a compact, modulardesign.
This cost-effective platform is ideallysuited for ISPs and enterprises that require innovative
universal or voicesen ices. The CiscoAS5350XM supports PSTN signaling, gateway
signaling, voicecodecs, fax. VoiceXML. RADIUS. Tool Command Language (Tel), and
interactive voiceresponse. The SBCfunctionality provides additional opportunities for IP-to-IP
trunking applications.
Cisco AS5400 Series Universal Gateway Platforms
The Cisco AS5400XM I 'nivcrsal Gateway offers unparalleled capacity in only 2 rack units
(Rl s) and provides data, voice, and faxseniccs. as well as SBCfunctionality. High-density,
low-power consumption, and a robust feature set make the Cisco AS5400XM Series Universal
Gateway ideal for several network deployment architectures, especially foreolocation
environments and for large points of presence (POPs). The Cisco AS5400XM Universal
(iateway offers reliable, scalable data and voice gateway functions and SBC services. The
Cisco AS5400XMsupports PSTNsignaling, gateway signaling, voice codecs, tax, VoiceXML.
RADIUS. Tel. and IVR.
Cisco 7200 Series Routers
Cisco 7200 Series Routers are sen ices routers for enterprise edge and service provider edge
applications. These compact routers provide serviceability and manageability coupled with
high-performance modular processors. The Cisco 7200 Series Routers offer a wide range of
gateway functions for voice, video. and data integration, and a comprehensive list of voice port
adapters that can be installed in the TDM-enabled VXR chassis. Due to its high performance,
this platform is well-suited to act as a Cisco Unified Border Element or H.323 gatekeeper.
1-32 Implementing Cisco Voice Communications and QoS (CVOICE) v8 0 2010 Cisco Systems. Inc.
Gateway Operational Modes
This topic describes the three operational modes ofvoice gateways.
Voice Gateway Overview
Three operational modes:
Voice switching
- Switches between multiple
traditional voice networks
VoIP gateway
- Converts between traditional
telephony and VoIP
Cisco Unified Border Element
IP-to-IP gateway
Converts parameters between
multiple VoIP networks
PBX
PSTN
Voice gateways can bedeployed inthree modes. Asingle gateway can operate in one mode or
inmultiple modes at the same time. These modes areas follows:
Avoice-switching gateway connects various analog and digital voice circuits. This
functionality isequivalent totheoperation of central office switches and PBXs in
traditional telephony.
AVoIP gateway connects thetraditional telephony network lo the IP network. It converts
the signaling and media transmission methods used onone side tothe other side. VoIP
gateways provide physical access for local analog and digital voice devices such as
telephones, faxmachines, keysets, and PBXs.
Cisco Unified Border Element interconnects two IP networks. It terminates the signaling
sessionsandeither passesthroughor terminates the mediachannels.
Voice Gateway Call Legs
This topic explains the concept ofacall leg and describes the call leg types associated with
eachof the three voicegateway operational modes.
1-34
Voice Gateway Call Let
Gateway calf
processing:
Connects incoming call
leg to outgoing call leg
Two major call leg types
are POTS and VoIP
Applies parameters to
both call legs
VoIP parameters are
negotiated.
Inbound
Outbound
Voice
Gateway
_Ji_
Incoming
Call Leg
Outgoing
Call Leg
Avoice call over a packet or traditional telephony network is segmented intodiscrete call legs.
\\ hen a gateway receives a call setup, it performs a routing decision and sends the call setup
request tothenext dev ice. Theincoming part of thecall is referred toas theincoming call leg
and theoutgoingpart of the call is referred to as the outgoingcall leg.
OnCisco 10S routers, the call legs areassociated with dial peers. One dial peer corresponds to
onecall leg. Acall leg is a logical connection between twogateways or between a gateway and
a telephony device. If thegateway receives or fonvards thecal! overananalog or digital voice
circuit, the corresponding call leg is referred lo as POTS. If the gateway receives or fonvards
the call overan IP interface, the correspondingcall leg is referred to as VoIP.
Ihe call legsare relev ant for call routing. Beforea gateway makes the call-routing decision, it
mustapply thesettings defined intheincoming call leg. Inthecaseof POTS incoming call
legs, these parameters define howthe gateway collects the dialed digits and optional
applications. In the case of VoIP incoming call legs, these parameters describe the voice
transmission methods, such as codec, voice activity detection (VAD). and dual tone
multifrequency (DTMF)-related features. Theseparameters must be successfully negotiated
between the local andpreceding gateway before thecall can be forwarded to the next gateway
in the path.
Implementing Cisco Voice Communications and QoS (CVOICE) v8 0 2010 Cisco Systems, Inc
Voice-Switching Gateway
This subtopic explains the call legs associated with the voice-switching gateway.
Voice-Switching Gateway
Call Legs
Voice-Switching Gateway:
* Signals calls
- Analog signaling
SS7, ISDN, QSIG
* Converts between:
- Signaling types
- Voice format (analog,
digital)
- Interface types (T1/E1,
FXO, FXS, E&M)
Uses plain old telephone
service (POTS) call legs
L
Voice-Switching
Gateway
tl_ 1
Outgoing
Call Leg:
POTS
Incoming
Call Leg:
POTS
Avoice-switching gateway has traditional telephony interlaces. Multiple call signaling
protocols exist, such as SS7, ISDN. QSignaling (QSIG), and the analog signaling methods,
including supervisory signaling (loop-start, ground-start, immediate-start, wink-start, delay-
start), address signaling (pulse, DTMF) and informational signaling. The voice-switching
gateway receives and fonvards the call setup request over analog ordigital voice circuits. The
gateway may have toconvert the call signaling and the voice fonnal when the call traverses the
gateway from one port to another. The incoming and the outgoing call legs are the POTS call
legs.
VoIP Gateway
This subtopic explains the call legs associated with the VoIP gateway.
Call Legs
' . IP
Originating
Gateway
I tl__
Terminating
Gateway
tt_ J
Call Leg 1 Call Leg 2 Call Leg 3
(POTS) (VoIP) (VoIP)
R1 Inbound R1 Outbound R2 Inbound
Call Leg 4
(POTS)
R2 Outbound
Ihe gateway provides translation between VoIP and non-VoIP networks such as the PSTN. It
converts the signaling and voicesignal between traditional telephony circuitsand the VoIP
transmission inan IPnetwork. One of thecall legs is a POTS call leg. while theotheris a VoIP
call leg.
Inthe figure, theoriginating gateway hasthe POTS incoming call legandthe VoIP outgoing
call leg. The VoIP terminating gateway has the VoIP incoming call legand the POTS outgoing
call leg. Bothgateway s must first successfully negotiate the VoIPparameters associated with
theirrespective outgoing andincoming call legs before theVoIP tenninating gateway can
fonvard the call to the destination PSTN network.
This subtopic explainsthe call legs associated with Cisco Unified Border Element.
Call Legs
Cisco Unified Border
Element:
Proxies signaling
May use different
IP
signaling on both
sides
Proxies or passes media
- May hide addresses of
media channels
Uses VoIP call legs
- Negotiates VoIP
parameters
Cisco Unified
Border Element
tl J
Incoming
Call Leg:
VoIP
Outgoing
Call Leg:
VoIP
Cisco Unified Border Element fonvards an incoming VoIP call as another, outgoing VoIP call.
It receives a call setup request, negotiates parameters, and forwards the call setup request to the
next gateway. The incomingsignaling protocol may differ fromthe outgoing signaling
protocol. When the call is successfully signaled end-to-end, Cisco Unified Border Element may-
cither proxy the media channelwhich is referred to as flow-around-orlet the media channel
pass through the gateway without any modificationwhichis referred to as flow-through. The
media proxy function is necessary when the VoIP traffic parameters of the incoming call leg
differ from the VoIP parameters of the outgoing call leg. When Cisco Unified Border Element
proxies the media channel, it changes the IP addresses of the media packets. This feature is
very useful for security or connectivity reasons. Both call legs of a Cisco Unified Border
Element are VoIP call legs.
) 2010 Cisco Systems. Inc. Introduction to Voice Gateways
Summary
This topic summarizes the key points that were discussed in this lesson.
References
Summary
Cisco Unified Communicationsarchitecture integrates IP
communications, mobile applications, customer care,
telepresence, conferencing, and messaging.
Cisco Unified Communications gateways connect voice-
enabled communication networks
Gateways are deployed in one of four modes: single-site,
multisite with centralized processing, multisite with distributed
processing, or clustering over the WAN.
The newest family of enterprise gateways, Cisco 2900 and
3900 Series Integrated Services Routers, offers rich unified
communications features.
The incoming and outgoing call leg describes the input and
output procedure for a call processed by the voice gateway.
for additional information, refer to these resources:
Cisco 2900 Series integrated Senices Routers:
http: www cNco.com 'goO'XHl
Cisco 3900 Series Integrated Services Routers:
http:1 www eisco.com;go.'3900
hitp:' www.cKeo.com'go'2 8t)0
hup:' www cisco.com go-3800
Cisco ATA 186:
http: .www.cisco.com go'ataUS')
Cisco AS5350 Universal Gateway :
http:1 www eis.co.com.'go.'asXoO
Cisco AS5-400 Series Universal (iateway platforms:
http:- www.cisco.com'go/as5-100
Cisco 7200 Series Routers:
http:;'w ww. cisco.com 'go/720(1
Cisco Unified Border Element:
http: www u-co.cum go'eiibe
1-38 Implementing Cisco Voice Communications and QoS (CVOICE) u8.0 )2010 Cisco Systems, Inc
Lesson 2
Examining Gateway Call
Routing and Call Legs
Overview
Aprimary function of the Cisco Unified Communications gateways istoroute calls. The
process ofcall routing includes the processing ofincoming and outgoing call legs. This lesson
describes how call legs arecreated when inbound and outbound dial peers arematched. It
provides details about thedial-peer matching process and explains thedirect inward dialing
(DID) feature.
Objectives
Upon completing this lesson, you will be able todescribe how gateways route calls and which
configuration elementsrelateto incoming and outgoingcall legs.
Describe the functionsof POTS, VoIP dial peers, and call legs as components of a simple
VoIP network
Explain howgateways route calls end-to-end
Describe how to configure POTS dial peers
Explain howtousedestination-pattern options toassociate a telephone number with a
givendial peer, and describethe number matching process
Describe how the router matches inbound dial peers
Describe how the router matches outbound dial peers
Describe howthedefault dial peer is used in a gateway,when it is employed, and which
default commands are used
Explain the DIDfeature, describethe differences betweentwo-stage and one-stagedialing,
and explain what the DID feature does
Gateway Call-Routing Components
This topic describes dial peers and explains how they are involved incall routing onCisco
Unified Communications satewavs.
Inbound and Outbound Dial Peers
Adial peer is an addressable call endpoint.
Dial peers establish logical connections between call legs to
complete an end-to-end call
* Agateway uses two dial peers for each call:
Inbound: matches the incoming call
Outbound: matches the call destination
Call Selup Message
inbound Dial Poor
Dial peers are essential to implementingdial plans and providingvoice seniccs over an IP
packet network. Dial peers are used to identify call source and destination endpoints and to
define the characteristics that tire applied to each call leg in the call connection.
A traditional voice call over the public switched telephone network (PSfN) uses a dedicated
64-kb/s end-lo-end circuit. In contrast, a voice call over the packet network is made up of
discrete segments or call legs. A call leg is a logical connection between two routers or between
a router and a telephony device. bach voice gateway establishes at least two call legs.
['he incoming call leg is associated with the inbound (source) dial peer, while the outgoing call
leg is associated with the outbound (destination) dial peer, as shown in the diagram in the
figure. Attributes that are defined in a dial peer are applied to that call leg.
Call legs are router-centric. When an inbound call arrives on a gateway, the gateway finds the
inbound dial peer and processes its settings. IIThe settings are acceptable, the gateway finds the
outbound dial peer, establishes the outgoing call leg. and the call is switched fromthe incoming
call leg to the outgoing call leg. You need to configure dial peers to enable call routing on a
gatew av.
1-40 Implementing Cisco Voice Communications and QoS (CVOICE] v8.0 )2010 Cisco Systems, Inc
This subtopic describes the various types ofdial peers.
Typeof Pal Peer [Network Technology
Plain old telephone
service (POTS)
VoIP
Multimedia Mail
overlP(MMolP)
Maps a dial stringto a specificvoice port on fte local
gateway. Thevoiceportconnectsthe gatewaytothe
PSTN, PBX, or analog telephone.
Points to the IP address or DNS name of the destination
VoIP device that terminates the call. This mapping
applies to VoIP protocolssuch as H.323or SIP.
Thedial peer is mappedtotie emailaddress ofthe
SMTPserver. This type of dial peer is used for
store-and-forward fax (on-ramp and off-rampfaxing).
Dial peers are generally classified into plain old telephone service (POTS) dial peers and
network dial peers.
POTS dial peers define the characteristics ofatraditional telephony network connection. The
POTS dial peer maps adial string to aspecific voice port on the local gateway. Normally, the
voice port connects the gateway tothe local PSTN, PBX, oranalog telephone.
The specific type ofnetwork dial peer used depends on the network transport technology. The
VoIP dial peer isby far the most common network type. The most prevalent network dial peers
are the following:
VoIP: The dial peer is mapped totheIPaddress, Domain Name System (DNS) name, or
server type of thedestination VoIP device thatterminates the call.
Multimedia Mailover IP (MMoIF): The dial peeris mapped totheemail address of the
Simple Mail Transfer Protocol (SMTP) server. This type ofdial peer isused for store-and-
forward fax (on-ramp and off-ramp faxing).
Dial Peers
This subtopic provides an overview of the two most common dial peers: POTS and VoIP.
POTS and VoIP Dial Peei
Telephone ^^^
1001 bfc/**^ ^
1/0/0 dUtiiwteq^dMth 11
Voice
Gateway
VoIP
session large! tpv4*l/2 16 1.1
destination-pattern 2Q01
Voice
Gateway
Telephone
2001
In the diagram, an analog telephone is connected to the Cisco Unified Communications
gateway. The gateway needs two dial peers. The POTS dial-peer configuration includes at least
the telephone number of theanalog telephone andthe voice portto which it is attached. Based
onthis information, the gateway forwards calls destined lo the defined telephone overthe
specified port.
The VoIP dial peer isconnected tothe IP network. The VoIP dial-peer configuration includes
at least the destination telephone number (or range of numbers) and thenext-hop IPaddress or
name used to progress the call further.
For call routing tosuccessfully forward calls inboth directions, at least these call-routing
elements are needed in every \oice-processing system:
Anappropriate POTS dial peerthatspecifies towhich voice portthetelephone is attached.
This applies only to the edge voice-processingsystems.
An appropriate VoIP dial peer that specifies the recipient destination address, or at least the
address of the next hop.
VoIP Dial Peers
This subtopic provides detailsabout the VoIPdial peers.
VoIP Dial Peer
* Points to either H.323 or SIP device
- MGCPconfiguration does not use voice network dial
peers
Sets the attributes of the network connection, such as:
- VoIP codec
- Capability to use voice activitydetection (VAD)
- Capabilityfor dual tone multifrequency (DTMF) relay
The dial-peerparameters varybasedon the dial-peertype. AVoIPdial peer can point to either
an 11.323or Session Initiation Protocol (SIP) device. A Media Gateway Control Protocol
(MGCP) device is not an option due lo its call agent-centric nature. When Cisco Unified
Communications Manageruses MGCP to control the voicegateway, the dial plan is maintained
and Cisco Unified Communications Manager makes the routing decisions. The gateway merely
receives instructions on how to process the voice circuits.
VoIP dial peer parameters include coder-decoder (codec), quality of service (QoS), voice
activity detection (VAD). dual tone multifrequency (DTMF) relay, and fax rate.
>2010 Cisco Systems. Inc. Introduction to Voice Gateways 1-43
VoIP Dial Peer Examples
This subtopic prov ides typical examples of VoIP dial peers.
VoIP Dial Peer Exampl
Cisco Unfed Communications
Manager Cluster
Voice Gateway
Voice-Mail Server
H.323 Gatekeeper
VoIP dial peers map a dial string to a remote network device. Some examples of these remote
network devices are as follows:
Cisco Unified Communications Manager cluster
Another voice gatew ay
SIP proxy
Voice-mail server
H.323 gatekeeper
1-44 Implementing Cisco Voice Communications and QoS (CVOICEI v8 0 2010 Cisco Systems. Inc
mm
<*
*te
End-to-End Call Routing
This topic describes the key differences between IP packet routing and call routing.
iP and Call Routing Comparison
wmm IP routina [Call routing H
Static or dynamic Onty static.
IP routing table Dial plan.
IP route Dial peer.
Hop-by-hop routing
each router makes an
independent decision
Inbound and outbound call legs. The gateway
negotiates VoIP parameters with preceding and
next gateways before a call is forwarded.
Destination-based routing Called number, matched by destination pattern, is
one of many selection criteria.
Longest-match rule The longest-match rule for destination pattern
exists but other criteria have higher priority.
Equal paths Preference can be applied to equal dial peers. Ifall
criteria are the same, random selection.
Defauit route Possible. Often points at external gateway or
gatekeeper.
Acomparison of IPpacket routingand call-routing principlesis helpful to understand the call-
routing process.
The entries that define where to forward calls are the dial peers. All dial peers together build
the dial plan, whichis equivalentto the IP routingtable. The dial peers are static in nature.
Hop-by-hopcall routing builds on the principle of call legs. Before a call-routing decision is
made, the gateway must identify the inbound dial peer and process its parameters. This process
may involve VoIP parameter negotiation.
The call-routing decision is the selection of the outbound dial peer. This selection is commonly
based on the called number when the destination-pattern command is used. The selection may
be based on other information, and that other criteria may have higher precedence than the
called number. When the called number is matched to find the outbound dial peer, the longest
match rule applies.
If more than one dial peer equally matches the dial string, all the matching dial peers are used
to forma so-called rotary group. The router attempts to place the outbound call leg using all the
dial peers in the rotary group until one is successful. The selection order within the group can
be influenced by configuring the preference.
A default call route can be configured using special characters when matching the number.
Call Routing
This subtopic introduces the taskof pathselection when routing calls
Call Routint
Multiple Paths
Minolta
Dial2001 2009K
^
1/0/0
Primary path
Secondary path, call forwarded to 300 555-2001
(requires digit manipulation for routing through PSTN)
2001
2002
2003
The VoIP gateway is often faced with the task of selecting the best path for a given destination
number. Such a requirement arises when the prelerred path goes through the IP WAN, and the
backup PSTN path should be chosen when the IP WAN is either unavailable or lacks the
needed bandwidth resources.
The figure illustrates a scenario with two locations connected to the IP WAN and PSTN. When
the call goes through the PSTN, its numbers (both calling and called) may have to be
manipulated so that they are reachable within the PSfN network. Otherwise, the PSTN
switches will not recognize the called number and the call will fail.
r~*
Call Legs
This subtopic explains the call legs on apair of VoIP gateways that participate in an cnd-to-cnd
call.
Call Legs
Source Gateway Perspective
ir
Dial2O01 RV 10.1.1.1
Inbound * l Outbound Call Leg
CaH L (VoIP Dial PeerinPrimary Path.
(POTS Dial POTS Dial Peer in Secondary Path)
Peer)
IP WAN
PSTN
^
2002
2003
The figure illustrates the call legs that are processed ona gateway that receives acall from a
locally attached telephone and originates aVoIP session. These call legs are created when the
telephone (1001) attached toan Rl gateway dials atelephone number in another location
(2001). When a call arrives onRl. the gateway creates aninbound call leg that corresponds to
the inbound dial peer, makes a routing decision by finding an outbound dial peer, and creates
anoutbound call legbyforwarding the call toward thedestination. Iftherouting decision
chooses an IPWAN. theoutbound call legwillbe VoIP; if therouting decision chooses a
PSTN, the outbound call leg will be POTS.
>2010 Cisco Systems. Inc.
Cal! Legs (Cont.)
Destination Gateway Perspective
Dial 2001 ri 101.1.1
InboundCall Leg +i Outbound
(VoIP Dial Peer in Primary Path. callLeg
POTS Dial Peer inSecondary Path) (POTS Dial
Peer)
IP WAN
PSTN
ir
2001
The figure illustrates the call legsthat are processed on the gateway that terminates the VoIP
session and fonvards the call to the locally attached telephone with extension 2001. The
inbound call legis createdwhenthe call arrives either through the IP WAN or the PSTN
network. Thegateway makes therouting decision byselecting theoutbound dial peer. The
outbound call legcorresponds to a POTS dial peer that points to the voiceport I/0/0. wherethe
recipient telephone is attached. Thegateway signals anincoming call on that port, andthe
telephone rings.
Implementing Cisco Voice Communications and QoS (CVOICE) v8 0 2010 Cisco Systems. Inc
This topic outline the objectives of call rooting through the PSTN network and desenbes the
corresponding configuration process.
Call Routing Through PSTN
Dial 2001
PSTN Path
(requires digit manipulation for routing through PSTN)
^
2001
2002
2003
This lesson explains how to configure the POTS dial peers that effectively enable call
fonvaSing along the PSTN path. The configuration of the primary VoIP path will be covered
in detail in alater lesson. The digit manipulation requirement is not covered at this time.
Configuring POTS Dial Peer*
Unidirectional Call Routing
=nfig) dial-peer voice 1
RKconfig-dialpe
Rl(config-dialpe
RUconf ig-dialpe
Dial 2001
on-pattern 2001
ill forward-digits
) port 1/1/C
R2(cc.nig)# dial-peer voice 1 pots
R2{config-dialpear)lf destination-patte
K2(config-dialpeerl# port 1/0/1
2001
Tli^figure illustrates ihe configuration that enables calling from extension 1001 to extension
The outbound dial-peer type is POTS because the destination number 2001 is reachable over a
1OISvoice port 1,1/0. Ihere are two basic parameters that need to be specified on this dial
peer: the telephone number and ihe voice port.
The destination-pattern command is used to match the called telephone number The Rl
gateway uses the destination pattern "2001".
The port command specifics the respective voice port. In this example, port 1/1/0 defines that
the port is on module 1. voice interface card (VIC) slot 1. and voice port 0.
The forward-digits all command makes the gateway send the entire called number in the call
signal to the next gateway. By default, the explicitly matched digits are discarded when the call
ts forwarded over an outbound POTS call leg. In this case, the destination pattern "2001"
matches explicitly all tour digits, so the gateway would not send any digits when forwarding
the call through the PSTN. The digit consumption rule applies only to outbound POTS call
legs. \\hen acall ,s forwarded over an outbound VoIP call leg, no digit consumption occurs bv
default, and all digits are sent to the next VoIP device.
Implementing Cisco Voice Communications and QoS (CVOICE! v8 0
Configuring POTS Dial
Bidirectional Call Routing
dial-peer vc ice 1 pots
deBtinotioc -pa Item 2001
forward-dig its all
port 1/1/0
dial-peer v= ice 2 pOtH
destination -pattern 1001
port 1/0/0
PSTN
2001
2/1/0
dial-paar voica 1 pots
deBtination-patfeam 1001
forward-digits all
port 2/1/0
dial-paer voice 2 pots
dastination-pattsrn 2001
port 1/0/1
The figure illustrates theconfiguration that enables call routing both ways. Additionally, the Rl
gateway has the POTS dial peer2 that matches extension 1001 and points tothevoice port
1/0/0. where thetelephone is attached. The R2gateway, inaddition to having the POTS dial
peerthat points totheattached telephone, has the POTS dial peer2, which matches the
extension 1001 and points toward the PSTN.
The forward-digits all command is usedon both gateways inthe dial peers pointingto the
PSfN. Without this command, the gateways woulddiscardthe explicitly matcheddigits when
sendingthe call to the PSTN. Nodigits wouldbe forwarded.
12010 Cisco Systems, Inc. Introduction to Voice Gateways
Dial Peer Matching
This topic liststhedial peer commands that matchtelephone numbers.
Use of String Matching
* Called number: Dialed Number Identification Service (DNIS)
"Calling number: Automatic Number Identification (ANI)
router(config-dialpeer] #
destination-pattern string
* Matches called number in outbound dial-peer
* Matches calling number in inbound dial-peer
router(config-dialpeer] #
I incoming called-number string
Matches called number in inbound dial-peer
router {conf ig- dialpeer) tt
answer-address string
Matches calling number in inbound dial-peer
When configuring dial peers on a Cisco Unified Coninuinications gateway, you can use three
commands that match telephone numbers. Two telephone numbers are usually sent with the
call: the calling number, known in ISDN as the Automatic Number Identification (ANI). and
the called number, referred to as the Dialed Number Identification Service (DNIS). Both
numbers can be used lo find the inbound and outbound dial-peer.
I he obvious usage of the destination-pattern command is to match the outbound dial-peer
based on die called number. The command is also considered when matching the inbound dial-
peer, but then the destination pattern string is matched against the calling number.
The incoming called-number command is only considered when selecting the inbound dial-
peer. It matches the original called number.
The answer-address command is only considered when selecting the inbound dial-peer. It
matches the original calling number.
1-52 Implementing Cisco Voice Communications and QoS (CVOICE) vB.O 2010 Cisco Systems. Inc.
This subtopic describes the regular expressions
that areused tomatch number strings.
Plus sign (+)
Period(.)
Percent sign (%)
Question mart (?)
CrcumttexO
Dollar sign ($)
T
Backslash (\)
Brackets 1]
Parentheses ()
As first character, indrates E.164 standard number; otherwise,
speeffles that the preceding dotoccurred one or more times
Matches any entered dial (used as a widcard)
Indicates that the preceding digit occurred zero ormore times
Indicates that the preceding digit occurred either zero orone time
Press Ctrt-vto disable context-sensibve help and enter ?character
Indicates a match tothebeginning ofthestring .
Matches the null string at the endofthe string
Timer character, indicates a variable-length delstring. Makes the
router wart until al digits arereceded before rouBng call
Followed by a single character, matches that character
Indicates a range
Indicates a pattern
The three string-matching commands, destination-pattern, incoming called-number. and
answer-address, have astring parameter. The gateway compares the received numbers with
the strings defined in the respective commands. The string may explicitly match the characters
in the telephone numbers (0-9. A-D. *. #), and it can contain special regular expressions:
. Plus sign (+): The plus sign in front of astring specifies that the string must conform to
E. 164. E.164 is the ITU-T recommendation for the international public telecommunication
numbering plan.
. Aperiod {.) matches any single entered digit from 0to 9, and is used as awildcard. The
wildcard can be used to specify agroup of numbers that may be accessible via asingle
path Apattern of "200." allows for 10 uniquely addressed devices, while apattern oi
"20 "can point to 100 devices. If one site has the numbers 2000 through 2049, and another
site has the numbers 2050 through 2099, the bracket notation would be more efficient.
. Brackets (I ]) indicate arange. Arange is asequence of characters that are enclosed in the
brackets. Onlv single numeric characters from 0to 9arc allowed in the range In the
previous example, the bracket notation could be used to specify exactly which range of
numbers is accessible through each dial peer. For example, the first site pattern would be
2010-41" and the second site pattern would be "20[5-91.'\ In both cases, aperiod is used
in the last digit position to represent any single digit from 0to 9. The bracket notation
offers much more flexibility in how numbers can be assigned.
1-53
ii . ,CHT-{ } aClCr iS inc,uded at thc end of the destination pattern the router
oolees dtaleddigtts until the interdigit timer expires (10 seconds, bv del ult) o1"
press the number terminate character (the default is #,. The timer character must be an
uppercase I.
Question mark ,?,: In string matching, this character indicates that the preceding character
occurred zero or one time Cisco IOS CI. however, uses this character to invoJcontext
Z ^ i P" ^ '[ 8nd ' ke>'S simultanl>-t0 Usable the context-sensitive help
and allow the question mark to be entered in the character string.
Note
An asterisk () and pound sign {#) are not considered special characters They appear on
standard touch-tone dial pads and may be used when passing acall to an automated
application that requires these characters to signal the use of aspecial feature For
example, when auser calls an interactive voice response (IVR) system that requires acode
for access, the number dialed might be "5551212888T, which would initially dial the
telephone number "5551212" and input acode of "886" followed by the pound key to
terminate the IVR input query
Implementing Cisco Voice Communications and QoS (CVOICE) V8 0 2010 Cisco Systems. Inc.
Number-Matching Characters
This subtopic presents examples of number matching.
Number-Matching Examples
mmm Nnmh-Rhino 1Matching telephone numbers
5551234 Matches single number 5551234
A5551234$ Matches single number 5551234
555123(5-9] Matches the numbers 5551235-5551239
55512[3-4j. Matches 7-digit numbers where the first 5 digits are 55512, the
sixth digit is 3 or 4, and the last digit is any digit
T Matches any number with the length of 1 to 32 digits
(200)75551234 Matches the numbers 2005551234 and 5551234
1(2-3]%4 Matches numbers that start with 1, have any number of
occurrences of the digit 2 or 3. and end with 4
The table provides examples of number matching:
String Matching Telephone Numbers
5551234 This pattern matches one telephone number exactly, 5551234.
This destination pattern is typically used when there is a single device, such as a
telephone or fax, connected to a voice port.
*5551234$ This pattern matches the number 5551234 using an explicit match of the beginning
and the end of the string.
555123(5-9] This pattern matches the number range 5551235 to 5551239.
55512[3-4], This destination pattern matches a 7-digit telephone number where the first 5 digits
are 55512, the sixth digit can be a 3 or 4, and the last digit can be any digit.
This destination pattern is used when telephone number ranges are assigned to
specific sites. In this example, the destination pattern is used in a small site that does
not need more than 30 numbers assigned.
T This destination pattern matches any telephone number that has at least 1 digit and
can vary in length from 1 to 32 digits.
This destination pattern is used for a dial peer that services a variable-length dial
plan for local, national, and international calls. It can also be used as a default
destination pattern so that any calls that do not match a more specific pattern will
match this pattern and can be directed to an operator.
(200)75551234 Matches the numbers 2005551234 and 5551234. This expression uses a pattern
(200) that can occur 0 or 1 time.
1[2-3]%4 Matches numbers that start with 1, have any number of occurrences of the digit 2 or
3, and end with 4. This expression uses a range [2-3] that can occur o or more
times
Matching Inbound Dial Peers
This topicdescribes the rules used by a Cisco Unified Communications gateway to matchan
inbound dial peer.
Matching inbound Dial Peers
Elements in the Call Setup Message
To match inbound call legs to dial peers, the gateway
uses one of three elements (ISDN example):
Called number (DNIS)
Derived from the ISDN setup message or channel
associated signaling (CAS) DNIS
Calling number (ANI)
Derived from the ISDN setup message or CASANI
- Inbound voice port
Call Setup Message " one
1
The inbound dial peer determines thecal! properties for the incoming side of the call. To match
inbound call legs to dial peers, the router uses three elements in the call setup message and live
configurable dial-peer attributes. The three call setup elements are, in the example of ISDN, as
follows:
Called number (OMS): Specifies the destination, which is derived from the ISDN setup
message or channel associated signaling (CAS) DNIS
Calling number (AM): Denotes the origin, which is derived from the ISDN setup
message or CAS AN!
Voice port: Carries die incoming call
1-56 Implementing Cisco Voice Communications and QoS (CVOICE) u8.0
Matching Inbound Dial Peers (Cont)
Relevant Dial-Peer Attributes
Precedence of matching criteria for inbound dial peers:
incoming called-number: Matches called number
- Most explicit match
',' answer-address: Matches calling number
Most explicit match
3 destination-pattern: Matches calling number
- Most explicit match
port. Matches the dial peer with the inbound voice port (POTS
only)
If multiple dial peers have the same port, selects the dial peer
added to the configuration earlier
: default dial peer: Predefined parameters
* Only one condition must be met.
The gateway stops searching when a dial-peer match is found.
The gateway selects an inbounddial peer by matchingthe informationelements in the setup
messagewiththe dial-peerattributes. The gateway matches these itemsin the following order:
1. Called number with incoming called-number
First, the gateway attempts to match the called number of the call setup request with the
configured incoming called-number parameterof eachdial peer. This attribute has
matching priority over the answer-address and destination-pattern matching. If multiple
incoming called-number attributes match the DNIS, the longest match wins.
2. Calling number with answer-address
If no match is found in Step 1, the gateway attempts to match the calling number of the call
setup request with the answer-address of each dial peer. This attribute may be useful in
situations where you want to match calls based on the calling number. If multiple answer-
address attributes match the ANI, the longest match wins.
3. Calling number with destination-pattern
If no match is found in Step 2, the gateway attempts to match the calling number of the call
setup request to the destination-pattern of each dial peer. If multiple destination-pattern
attributes match the DNIS, the longest match wins.
4. Voice port (associated with the incoming call setup request) with the configured dial peer
port parameter (applicable for inbound POTS call legs)
If no match is found in Step 3, the gateway attempts to match the configured dial-peer port
parameter lo the voice port associated with the incoming call. If multiple dial peers have
the same port configured, the dial peer first added in the configuration is matched.
5. If there is no match, the default dial peer is used. The default dial peer is explained later in
this lesson.
Only one condition must be met. The gateway stops searching when a dial-peer match is found.
>2010 Cisco Systems. Inc Introduction to Voice Gateways
Matching Inbound Dial
Example
P .., PSTN
dial-pear voice 1 pots
destination-pattern 20C
forward-digits all
pore i/i/O
Which inbound dial peer
is selected'
(answer-address)
2/1/0
dial peer vo ice 1 pots
destination -pa ctern 2001
port 1/0/0
dial peer vo ice 2 pots
ansv er-addrCBS 100.
port 2/1/0
dial peer vo ice 3 pcta
inc. ming enlled-number 100.
por 2/1/0
dial peer vo ice 4 pota
destination -pattern 100.
ton ard-dig its all
por 2/1/0
The figure illustrates an example of matching inbound dial peers. When the destination
gatewa; receives the call setup request, it looks for the inbound dial peer. The ANI is 1001: the
DNIS is 2001. The incoming called-number command has the first precedence and exists in
dial peer 3 but does not match the DNIS. The answer-address command has the second
precedence, exists in dial peer 2. and matches the ANI. Therefore, dial peer 2 is selected as the
incoming dial peer.
mi
Matching Inbound Dial Peers (Cont.)
Guidelines
answer-address
- Useful for matchingthe geographical regionof caller
Callers from a givencountrydirected at the appropriate
language-speaking team
Callers froma specific region directed at the regional
sales staff
incoming called-number
- Recommended for most configurations
- Useful for service selection
Different numbers for sales and technical support
- Differentnumbers for shipping order, tracking, and
cancellation
Usethe answer-address command whenmatchingthe geographical regionof the caller. This
approach is recommended in thesesituations:
Callers from a given country should be directed to theappropriate language-speaking team.
Callers from a specific region should bedirected tothe regional salesstaff.
Usethe incoming called-number command wheneverpossible. Because all types of call setup
messages and signalsalways include the DNISinformation, Ciscorecommends usingthe
incomingcalled-numbercommand for inbound dial peer matching. Inparticular, the
incoming called-number command is useful for serviceselection, such as in thesesituations:
Different numbers are available to reach the sales and technical support.
DitTerent numbers exist for shipping order, tracking, and cancellation.
Matching Outbound Dial Peers
This topic describes the rules used bya Cisco Unified Communications gateway tomatch an
outbound dial peer.
Matching Outbound Dial Peers
Criteria required for outbound dial peers:
' destination-pattern
Uses the called number to match the outbound dial peer
Most explicit match rule applies
Whena call setup request arrives on a voice gateway, the gateway uses the incomingdial string
to match the destination pattern in the outbound dial peer. Both dial peersPOTS and VoIP
are considered together for outbound dial peer matching.
Once the outbound dial peer is found, the call setup is progressed lo the next device along the
path. On outbound POTS dial peers, the port command is used to forward the call. On
outbound VoIP dial peers, the session target command is used to forward the call.
1-60 Implementing Cisco Voice Communications and QoS (CVOICE) vS.O >2010 Cisco Systems. Inc
Matching Outbound Dial Peers (Cont.
Example
dial-peer voiCB 1 pots
destination-pattern ,T
port 1/1/0
dial-poor voice 2 pots
destination-pattern 20[0-11.
forward-digits all
port 1/1/0
destination-pattern 200.
forward-digits all
port 1/1/0
dial-peer voice * pots
destination-pattern 2001
forward-digits all
port 1/1/0
Dialed number 2001 matches dial peer 4.
Dialed number 2002 matcties dial peer 3.
Dialed number 2011 matcties dial peer 2
Dialed number 2111 matches dial peer 1
2001
2O02
2011
2111
The figure illustrates an example of outbound dial peer matching. Four calls are made from the
telephone withextension1001:
The userdials 2001. Thebestmatch is found withdial peer4.
The user dials 2002. Dial peer 1matches that number, but the match isthe least specific.
Dial peer 2matches that number and also atotal of20 numbers (2000 to 2019). Dial peer 3
matches that number and also atotal of 10 numbers (2000 to2009). Dial peer 3yields the
best match.
The user dials 2011. Dial peers I and 2match the number, with the latter offering the
longest match.
The user dials 2111. Only dial peer 1 matches.
i 2010 Cisco Systems. Inc.
Default Dial Peer
This topic describes the default dial peer and its usage when no explicit dial peer ean be
matched.
Default Dial Peer
No explicit match for inbound dial peer configured
dial-peer vo ice 1 pots
destination -pa teem 200.
orard-dig its all
port 1/1/0
Dial 2001
1/0/0 1/1/0 -RSTN-
R1
What are the inbound dial peers
when extension 1001 calls 2001?
2001
2002
2/1/0
2003
dial-peer v ics 1 pots
destinatio -pattern 2001
port 1/0/0
Ihe ligure illustrates the situation inwhich thecal! routing works only inonedirection. This
scenario brings up the question about the inbound dial peers selected on both gateways.
1-62 Implementing Cisco Voice Communications and OoS (CVOICE) v8.0
Default Dial Peer (Cont.)
Features
When no explicit inbounddial peer is configured, the
gateway uses the default dial peer:
Also called dial peer 0
Fails to negotiate any nondefault parameters
Inbound VoIP dial peer 0
psfaneters
G.729orG.711 codec
No DTMF relay
IP precedence 0
VAD enabled
NoRSVP support
Fax-rate voice
in&ound POTS dial peer 0
parameters
No applications
No DID
Ifnoinbound peer can bematched by the defined criteria, the gateway resorts tothe default
dial peer. The default dial peer isreferred toasdial peer 0. Default dial peers are used for
inbound matches only. They never match outbound calls. Thecharacteristics of dial peer0
cannot be changed.
Dial peer 0 for inbound VoIPpeers has this configuration:
G.729 and G.711 codecs are supported.
IP precedence is set to 0.
VAD is enabled.
RSVP is not supported.
Fax-rate service is supported.
Dial peer 0 for inboundPOTS peers has this configuration:
No applications
No direct inward dialing
Youcannot changethe default configuration for dial peer 0. Defaultdial peer 0 fails lo
negotiate nondefault capabilities, services, andapplications, suchas DTMF relay or disabled
VAD.
Whenthe default dial peer is matched on an inbound POTS call leg, there is no default IVR
application enabled ontheport. Asa result, the usergetsa dial tone andproceeds to dial digits.
Defauft Dial Peer (Conf.)
Guidelines
Avoid using dial peer 0
Calls with nondefault parameters will fail.
Incoming called-number ensures a match with the
desired parameters.
Many errors are due tocodec, VAD, and DTMF-relay
misconfigurations when dial peer 0 is matched.
Cisco AS5350, AS5400, andAS5850 Universal Gateways
require explicit inbound dial peers matching:
To accept incoming POTS calls as voice calls.
Ifthere is no inbound dial peer match, the call is treated
and processed as a dialup (modem) call.
Avoid using dial peer 0. Having theincoming called-number parameter configured correctly
ensures that thedial peer is alwav s matched with theparameters thatyouwant when placing
outbound calls through a gatewa>, Many problems with calling out through a Cisco IOS
gateway aredueto codec. VAD. and DIMF-relay misconfigurations when dial peer0 is being
matched.
When theCisco AS5350. AS5400. or AS5850 Iiniversal Gateway platforms do notexplicitly
match an incoming dial peer, dial peer 0 is matched and the call is treated as a dial modem call.
This call treatment can result in getting modemtones rather than a dial tone for inbound calls.
The explicitinbound dial peer matching on theseplatforms matches onlythe first three criteria
(incoming called number, answer address, destination pattern) andignores the incoming port
information. Therefore, if the incoming called-number. answer-address, and destination-
pattern commands do not match, the call is treated as a modem call.
1-6d Implemenling Cisco Voice Communications and QoS (CVOICE) v8 0 2010 Cisco Systems. !nc
Direct Inward Dialing
This topic describes DID and explains the differences between one-stage dialing and two-stage
dialing.
Direct Inward Dialing
Two-Stage Dialingand One-Stage Dialing
Two-stage dialing
- When a call arrives on a POTS voice port
POTS voice port is seized inbound
Gateway presents dial tone and collects digits
One-stage dialing
- When a call arrives on a DID-enabled voice port
Gateway does not present a dial tone
Gateway receives the entire called number
Enabled through DID on inbound POTS dial peers
DIDnot supported on FXS/FXO/E&Manalog ports
DIDavailable on FXS-DID and digital circuits
In the early days of traditional telephony, enterprises used two-stage dialing to allowoutside
callers to reach internal telephones. An enterprise PBX was connected to the PSTN over an
analogor digital trunk. Whenthat trunkreceived an inbound call, the centraloffice (CO) switch
seized the voice port. The PBXpresented a dial tone and started collecting digits. The caller
heard a secondary dial tone fromthe enterprise PBXand dialed the number required to reach
the internal telephone.
With the invention of DID in the 1970s, one-stage dialing was made possible. With one-stage
dialing, the callersenter the entirecalledpartynumber, including the numberrequiredto reach
the internal telephone. They do not hear a secondary dial tone. The PSTNCO switch sends the
entire DNIS to the PBX. which forwards the call to the internal telephone.
Voice gateways can use DID if it is enabled on inbound POTS dial peers. It is supported on all
digital voice ports and the analog FXS-DIDports. It is not supported on analog Foreign
Exchange Station (FXS). Foreign ExchangeOffice (FXO), and ear and mouth (E&M) voice
ports.
) 2010 Cisco Systems, Inc Introduction to Voice Gateways
Two-Stage Dialing
This subtopic describes the process of two-stage dialing.
Overview
User dials 555
/ PSTN delivers call, gateway sends dial tone and starts
collecting remaining digits
User hears secondary dial tone and dials 2001
Gateway matches the outbound dial peer and signals the call
2001
dial-pser vc ioe 1 pots
port 1/1/1
The tlgure illustrates die process of two-stage dialing.
This process is depicted b\ the following steps:
1. The user takes the phone off-hook, recch es dial lone, and dials 555.
2. The PSTN receives the digits and delivers to the destination gateway. The trunk line to the
destination gaiewa\ is seized by the adjacent CO switch. The destination gateway presents
the secondary dial tone and starts collecting digits until it can identify an outbound dial
peer. Whether the digits are dialed with irregular intervals by humans or in a regular
fashion by telephonj equipment that sends the preeolleeled digits, dial-peer matching is
done digit-b>-digit. This means that the gateway attempts lo match a dial peer after each
digit is received.
3. The user hears the secondary dial tone and dials 200!.
4. The gatewa> uses the number 2001 to match the outbound dial peer.
The destination gateway signals an incoming call to the telephone on port 1/1/1. and the
telephone rings.
Implementing Cisco Voice Communications and QoS (CVOICE) v8 0 i 2010 Cisco Systems, Inc
Two-Stage Dialing (Cont.)
Digit-by-Digit Collection
Digitsare collected in-band
Outbounddial peer matchingdone on a digit-by-digit basis
Gateway matchesdial peer after receiving each digit
- Routing when a matchis found 2002-2009
Dial 555-2001
dial-peer voiaa 1 pots
de Btination-pattern
2001
port 1/1/1
port 1/1/0
Ihe figure illustrates a potential issue related totwo-stage dialing. Thedestination gateway
uses an incorrectly designed dial pian. Because the destination gateway collects the dialed
digits in-band. onadigit-by-digit basis, it matches dial peer 2before the complete number has
been received. The call cannot be delivered to its intended recipient.
Wildcard Use
* Most explicit match rule applies
Destination gateway matches dial peer 1 for the outbound
call leg
Dial 555-2001
dial-peer vt ice 1 po
destinatior -patt rn 2001
port 1/1/1
dial-peer v; ice 2 po s
defltinatior -pattern 200.
port 1/1/0
2001
Tosolve the problem of theincorrect Kdesigned dial plan andtwo-stage dialing, youshould
use a wildcard inthedestination pattern of thedial peer2. 'fhis causes thedestination gateway
to wait for four digits before making the call-routing decision. With this solution, the call can
be deli\ ered to its intended recipient.
Gateway-by-Gateway Processing
ooo
Dial 55 4 2001
dial-peer voice 1 pots
port 1/0/1
...Irregular interval a-
0 0 0 0 0 0-0
..IicvguldE IntarVila
*>2> 0 0 0 0 0
dial-peer voice 1 pots,
destination-pattern A
port 1/0/1
...Irregular interval*.. R3
0 0"00
<C
dentinalion-pattern 2001
port 1/0/1
1/0/1
2001
The figure illustrates the process of two-stage dialing ona gateway-by-gateway basis:
1. The usertakesthephone off-hook andreceives thedial tonefrom local gateway Rl.
2. The user dials 55 by enteringthe digits in irregular intervals.
3. Rl collects thetwodigits (55), matches theoutbound dial peer, andseizes thetrunk line
1/0/1 toward R2.
4. R2 presents the second dial tone.
5. The user hears the secondary dial tone and dials 4.
6. R2 matches the outbound dial peer. R2 seizes the trunk line 1/0/1 to R3.
7. R3 presents the third dial tone.
8. "Iheuser hears the third dial tone and dials 2001 by entering the digits in irregular intervals.
9. R3 keepscollecting digits until the number2001 has beenreceived. That numbermatches
the outbound dial peer.
10. R3signalsan incoming call to the voiceport 1/0/1. The recipientphonerings.
2010 Cisco Systems, Inc Introduction to Voice Gateways 1-69
Variable-Length Numbers
The gateway waits for remaining digits.
The default interdigit timeout is 10 seconds.
The interdigit timeout can be modified using the timeouts
interdigit command in voice-port configuration mode.
Dial 555 2001
PSTN
Incoming called-number
I
destination-pattern .T
port 1/0/0
0l" 0^0 00 0
Thereare situations in which expected dial stringsdo not havea set numberof digits. Insuch
cases, it is usually best to use variable-lengthdial peers by configuring the T terminator on the
dial-peer destination-pattern command. When the timer (T) character is included at the end
of thedestination pattern, therouter collects dialed digits until theinterdigit timer expires (10
seconds. b\ default) or until the termination character (the default is #) is dialed.
1-70 Implementing Cisco Voice Communications and QoS (CVOICE) v8 0 J2010Cisco Systems. Inc
One-Stage Dialing
This subtopic explains DID and how one-stage dialing differs from two-stage dialing.
One-Stage Dialing
DID Overview
User dials 555-2001
PSTN sends the entire called number in one call setup
message to destination gateway (ifdigital circuit)
Destination gateway matches the outbound dial peer and
signals the call
Dial 555-2001
direct-inward-dial
port 1/1/1
One-stage dialingis enabledwhen the DIDfeature is configured on the inboundPOTS dial
peer of the destination voicegateway. With one-stage dialing,the destination gatewaydoes not
present thedial tone. Therefore, thecallerenters theentirenumber without hearing any
secondary dial tone. The PSTN can deliver the callednumberto the destination gateway intwo
ways:
Over digital interfaces: The COswitchsends a call setupmessagethat containstheentire
DNIS. The DNIS is mapped to the outbound dial peer. The gateway forwards the call
direct!} to the configured destination.
Over analog interfaces (FXS-DID): The digits are automatically signaled to the
destination gateway by the switch, without the requirement of the secondary dial tone.
The figure illustrates one-stage dialing:
1. Ihe user takes the phone off-hook, receives the dial tone, and dials 555-2001.
2. The PSTN delivers the call to the destination gateway. The destination gateway receives
the last four digits of the called number in one call setup message or over an analog FXS-
DID trunk.
3. The destination gateway matches the outbound dial peer and signals an incoming call to
port 1/1/1. The recipient phone rings.
i 2010 Cisco Systems, Inc. Introduction to Voice Gateways
One-Stage Dialing (Cont)
Matching with Complete Called Number
With DIDconfigured in the inbound POTS dial peer, the
router uses the complete called number to match the
outbound dial peer n
K 2002 2009
Dial 555-2001
-3&
PSTN
(-<$?
dial-peer v nice 1 pota
incoming e illed-number .
direct-inw ird-dial
dial-peer v >ice 2 pots
destinatjo i-pattern 2001
port 1/1/1
dial-peer v iice 3 pots
port 1/1/0
Ihe figure illustrates how DID manages the problem of incorrectly designed dial plans. The
destination gatewa> is connected to the PSTN o\era digital trunk and has the DID feature
enabled on the inbound POT'Sdial peer. Because the PS'I'N has consumed the tirst digits from
the called number, the destination gateway receives the lasl four digits in one call setup
message. 1 he destination gateway selects dial peer 2 as the best match. The call reaches the
intended recipient.
1-72 Implementing Cisco Voice Communications and QoS (CVOICE) v8 0 2010 Cisco Systems, Inc
One-Stage Dialing (Cont)
Gateway-by-Gateway Processing
dial-peer voica 1 pots
port 1/0/1:0
The figure illustrates theprocess of one-stage dialing on a gateway-by-gateway basis:
1. The user takes the phone off-hook and receives the dial tone fromthe local gateway Rl.
2. The user dials 554-2001 by entering the digits in irregular intervals.
3. Rl collects the number, matches the outbound dial peer, and finds tliat the outgoing call leg
goes over the digital trunk 1/0/1:0. Because the outgoingvoiceport is a digital circuit, Rl
sends the entire called number in one call setup message. Rl does not forward the first two
digits(55) becausethey are explicitly matchedby the destination patternand are consumed
by default (forward-digits all is not configured).
4. R2 receives the called number (4-2001), matches the outbound dial peer I, and forwards
the called number (2001) in a single call setup message over the outgoing digital trunk, to
R3. The first digit (4) is consumed throughtheexplicit matchby the destination pattern.
5. R3 receives the called number (2001) in a single message and matches the outbound dial
peer.
6. R3 signals an incoming call to port 1/0/1 and the recipient phone rings.
Note Ifanalog trunks were used in this example, the digits would be sent sequentially. The caller
would not hear any secondary or tertiary dial tones.
) 2010 Cisco Systems. Inc Introduction to Voice Gateways 1-73
One-Stage Dialing (Conl
Configuring DID
PSTN
String Description
dial-peer v >ice 1 pots
incoming c illed- umber .
direct-inw td-dial
dial-peer v ice 1 pots
deatinatio -pattern .T
port 1/0/0
Matches any number with at least one digit. Useful for outbound
(destination-pattern) and especially Inbound matching (incoming called-
number).
Matches any number with at least one digit. The timer character matches
either the interdigit timeout or the termination character (#). Useful for
outbound matching (destination-pattern).
I he DID feature is configured using the direct-inward-dial command in the incoming dial
peer. The inbound dial peer can be matched in various ways. The recommended method to
match inbound dial peers is to use the incoming called-number command. The figure displays
two most common]} found DID configurations, while the method using the incoming called-
number command is preferred. Note that the Time character (T) is not used in the incoming
called-number command, although it is used in destination patterns.
The following is the explanation of the two strings:
The string . (single period) matches any number with at least one digit. It is useful for
outbound matching using the destination-pattern command, and especially for inbound
matching \\ ith the incoming called-number command.
The string .1 (single period followed by '[') matches any number with at least one digit. The
timer character matches either the interdigit timeout or the termination character (#). The
string is useful for outbound matching using the destination-pattern command.
1-74 Implementing Cisco Voice Communications and QoS (CVOICE) vB.O )2010 Cisco Systems, Inc.
Summary
This lopic summarizes the key points that were discussed in this lesson.
Summary
The most common dial peer types are VoIPand POTS.
In call routing, each gateway identifies the inbound and
outbound dial peer.
POTS dial peers facilitate calling over POTS ports.
Telephone numbersare matchedusinga sequence of
standard and special characters.
The matching orderfor inbound dial peer is: incoming dialed-
number, answer-address, destination-pattern, and port.
The outbounddial peer is found by using the longest match
of the destination-pattern command.
Ifno explicit inbound dial peer is identified, the default peer 0
is used to set the parameters to predefined values.
DID enables the matching of the entire number instead of
digit-by-digit matching.
Lesson 3
Configuring Gateway Voice
Ports
Overview
Connecting voice devices to a networkinfrastructure requiresan in-depth understanding of the
signalingand eharactcri stiesthat are specificto eachtype of interface. Digital trunks arc used
to connectto the publicswitched telephone network (PSTN), to a PBX, or to the WAN, andare
widelyavailable worldwide. This lessonmapsout analoganddigital interfaces; examines
analog voice ports, analog signaling, and configuration parameters for analog voice ports; and
explains how to implement and verify digital trunks.
Objectives
Upon completing this lesson, you will be able to describe how to connect a gateway to
traditional voice circuits using analog and digital interfaces. This ability includes being able lo
Position the various types of analog and digital voice port interfaces in enterprise scenarios
Describe the various types of analog voice ports and their characteristics
Configure the analog voice ports
List the types of digital voice ports and describe their major characteristics
Describe ISDN and ISDN signaling
Configure Tl and El trunks to the PSTN
Configure ISDN PRI and BRI trunks
Tune various timers and parameters on analog and digital voice ports
Explain and configure echo cancellation
Verify analog and digital voice port configuration
Voice Ports Overview
This topic describes the ditTerent voice ports available on Cisco UnifiedCommunications
gateways and their deployment.
Voice Port Overview
Connecting End User Equipment
+*L
-y
^
>*
-*y
**
*>A
-^
^
FXS
(Analog)
Tl orEtor
'SON (Digrtali
Voice Port Vcuce Fort
FXS
(Ariatot
\ FXO
(Analog)
Voice ports on gatewavs emulate physical telephony switch connections, so that voice calls and
their associated signaling can be transformed between a packet network and a circuit-switched
network. Io make a voice call, certain information, such as the telephone on-hook status, the
availability of the line, and whether an ineoming eall is trying to reach a device, must be passed
between the end telephony devices. This information is referred lo as signaling, and to process
it correctly, the directly connected devices must use the same type of signaling.
Circuit-switched signaling is accomplished by installing appropriate voice hardware in the
router or access server and by configuring (he voice ports that connect lo telephony devices or
to the circuit-switched network.
The figure shows how traditional end-user equipment is commonly connected lo the PSTN.
Analog telephones are connected to the analog Foreign Exchange Station (FXS) interface
installed on the gateway . The gateway can then provide PSTN connectivity via either digital
circuits (II. El. and ISDN), or an analog Foreign Exchange Office (FXO) interface. Digital
circuits support man; connections over the same port. Analog ports support only one call per
port. The FXO interface can be deployed in tandem with an FXS-direct inward dialing (DID)
port, which would increase the number of simultaneous connections to two. In that case, the
FXS-DID port prov ides inbound connectivity while the FXO supports outbound calls.
Implementing Cisco Voice Communications and QoS (CVOICE) v8.0 ) 2010 Cisco Systems. Inc
This subtopic explains how PBXs are connected tothe PSTN.
Voice Port Overview (Cont.)
PBX Trunks
T1. E1,or ISDN
(Digital)
Voice Port Voice Port
FXO (Analog) *
additional FXS-DID
possible
Voice Port
T1.E1.W ISDN ^
(Digital) '
PSTN
PSTN
The figure shows the typical PBX connections. Most commonly, digital circuits are used to
carry many simultaneous calls over one voice port. Digital signaling interfaces include Tl, El.
and ISDN. Digital circuits can beenabled for QSignaling (QSIG) toexchange an extended set
ofprivate PBX features. Other connectivity options include the analog voice ports: FXO and
ear and mouth (E&M). It is important toknow which signaling method thatthe telephony side
of the connection isusing, and tomatch the router configuration and voice interface hardware
to that signaling method.
This subtopic describes Centralized Automated Message Accounting (CAMA).
Voice Port Overview (Cont.)
Centralized Automated Message Accounting Trunks
1
t-
y
Tl PRI lor Standard Cais
PSTN
CAMA Trunk
for EmergerKy
Calls
J*1
Ci-'.i~ '~*^H Public Safety
MU mat Answering Point
Centralized Automated Message Accounting (CAMA) available in Morth America only
ACAMA trunk isa special analog trunk type that is now mainly used for emergency call
services. Use CAMA ports toconnect toa public safely answering point (PSAP) for emergency
calls. ACAMA trunk canonly send outbound Automatic Number Identification (ANI)
information (callingnumber), which is required by the local PSAP. CAMAis available in
North America only.
Ihe calling number is needed at the PSAP for two reasons:
The catling number is used to reference a database to find the exact location of the caller
and an; extra information about the caller.
The calling number is used as a callback number in case that the call is disconnected.
The figure shows a voice gateway connecting an enterprise lo an Enhanced911 (E911)
network. Callsto emergency services are routed basedon the callingnumberrather thanthe
called number. The calling number is checked against a database of emergency service
providers that cross-references the service prov iders for the caller location. When this
information is determined, the call is thenroutedto the proper PSAP. whichdispatches serv ices
to the caller location.
Voice Trunk Example
Thissubtopic provides anoverview of howtrunks are used inanenterprise environment.
Voice Trunk Example
Chicago
"CAS = Channel assodated signaling
"CCS = Common channel signaling
A voice trunk (also called tie-line) is an analog or digital voice circuit that connects a gateway
to the PS'I'N. PBX. or another gateway. Lines connecting end-user devices are not called
trunks.
Trunk ports can be analog or digital and use various signaling protocols. Signaling can be done
using either the voice channel (in-band) or an extra dedicated channel (out-of-band). The
available features depend on the signaling protocol in use between the devices.
The diagram in the figure illustrates various possible trunk connections:
If a subscriber at the London site places a call to the PSTN, the gateway uses one voice
channel of the El R2 trunk interface.
If a subscriber of the legacy PBX system at the Chicago site needs to place a call to a
subscriber with an IP phone that is connected to the Chicago gateway, the call will route
via one channel of the E&M trunk between the legacy PBX and the gateway.
The Denver and the Chicago sites are connected to San Jose via QSIG to build up a
common private numbering plan between those sites. Because the Cisco IP telephony
introduction in Denver has not started yet, the QSIG trunk is established directly between
the San Jose gateway and the Denver legacy PBX.
Installing Voice Ports
This subtopic describes how voice ports are installed in Cisco router chassis.
^tailing Voice Ports
Cisco 2921/2951 Series router rear panel
"fhis figure illustrates the rear panel of ihe Cisco 2921/2951 Series router, a diagram of a voice
network module and of a voice interface card. The router includes two slots for Service
Modules and four slots for Enhanced High Speed WAN Interface Cards (EHWIC). The number
of supported modules and interface cards varies depending on the router platform. The slots are
typically numbered from right to left (when viewed from the rear) and from bottom to lop.
This figure depicts how the identifier of the voice port in the router configuration relates to the
physical voice port position in the router. In this example, the voice-port 2/0/1 command
indicates these parameters:
Voice port is installed in service module or network module inserted in module slot 2
The port interface card is installed in interface card slot 0 of that module.
The given port is the second interface from the right.
1-82 Implementing Cisco VoiceCommunications and QoS (CVOICE] v8.0
>
Analog Voice Ports
fhis topic describes thetypes and characteristics of analog voice ports.
Analog Voice Ports
FXS
FXS: Connects directly to end-user equipment such as
telephones, tax machines, or modems
FXO iFXO
^
PSTN
FXO: Used for trunk, or tie-line, connections to a PSTN CO or to
a PBX that does not support E&Msignaling
, E&M !?fe
E&M: Used for trunk circuits to connect telephone switches to
each other
Analog voice port interfaces connect routers in packet-based networks to analog two-wire or
four-wire analog circuits in telephony networks. There are three types of analog voice
interfaces that Cisco gateways support:
FXS interfaces: An FXS interface connects the router or access server to end-user
equipment such as telephones, fax machines, or modems. The FXS interface supplies ring,
voltage, and dial tone to the station and includes an RJ-11 connector for basic telephone
equipment, key sets, and PBXs.
FXO Interfaces: An FXO interface is used for trunk connections to a PSTN central office
(CO) or to a PBX lhat does not support E&M signaling. This interface is of value for off-
premises station applications. A standard RJ-11 modular telephone cable connects the FXO
voice interface card to the PS'I'Nor PBX through a telephone wall outlet.
K&M Interfaces: Trunk circuits connect telephone switches to each other. They do not
connect end-user equipment to the network. The most common form of analog trunk circuit
is the E&M interface, which uses special signaling paths that arc separate from the trunk
audio path to convey information about the calls. The signaling paths are known as the
E-lead and the M-lead. E&M connections from routers to telephone switches or to PBXs
are preferable to FXS and FXO connections because E&M provides better answer and
disconnect supervision.
The name E&M derives from the names of the two signaling leads, car and mouth, bul the
name "recEive and transMit" is also common.
Like a serial port, an E&M interface has a DTE or DCE type of reference. In the
telecommunications world, the trunking side is like the DCF and is usually associated with CO
functionality. The router acts as this side of the interface. The other side is referred to as the
signaling side, like a DTE. and is usually a device such as a PBX.
Analog Signaling Overview
This subtopic explains the analog signaling methods.
Analog Signaling Overview
* Supervisory signaling forFXO/FXS
Loop-start
Ground-start
Address signaling
Pulse
Dual tone multifrequency
Informational signaling
Call progress tones
Voice ports on routers and access ->er\ers physically connect the router or access server to
telephony devices such as telephones, fax machines. PBXs. and PSTN CO switches. These
devices may use any of several types of signaling inlerfaces to generate information about on-
hook status, ringing, and line seizure.
Signaling techniques can be placed into one of three categories:
Supervisory: Involves the detection of changes to the status of a loop or trunk. Once these
changes are detected, the supervisory circuit generates a predetermined response. FXO and
FXS interfaces indicate on- or off-hook status and the seizure of telephone lines by one of
two access-signaling methods: loop-start or ground-start.
Addressing: Involves passing dialed digits (pulsed or tone) lo a PBX or CO. These dialed
digits provide the switch with a connection path to another phone or customer premises
equipment (CPF).
Informational: Provides audible tones, which indicate certain conditions such as an
ineoming call or a busy phone, to ihe user.
iplementmg Cisco Voice Communicationsand QoS (CVOICE) v8.0
mt
<*
Analog Signaling
This subtopic compares the supervisory signaling on the analog voice ports.
Analog Signaling
Supervisory Signaling Compar son
^M
FXS / FXO
E&M
Wiring
RJ-11,two voice wires, 8-pin modular, up to S wires in
used for both total: 2 or 4 voice wires {half or
supervisory signaling fullduplex audio path) + 2 half or
and voice modulation full duplex wires for supervisory
signelingfM-iead, E-lead) + 2 for
voice modulation
Types of Loop-start (more Type I, II, III, IV. V, SSDC5
supervisory
common), ground-start
signaling
Supervisory Differences in voltage Differences in vottage and
signaling and grounding on the tip grounding on the M-tead and E-
differences and ring lines lead
Access signaling N/A
Immediate-start wink-start, delay-
start followed by pulse or DTMF
rI*he figure summarizes the major wiring and supervisory signaling differences between the
FXS and FXO. and the E&M voice ports.
TheFXS and FXO voice ports use standard RJ-11 modular telephone cable and use two wires
to connect to the adjacent telephony device.
The phy sical E&M interface isan eight-pin modular connector that connects toPBX trunk
lines, which are classified as either two- or four-wire. This refers to whether the audio path is
full duplex onone pair of wires (two-wire) orontwo pair of wires (four-wire). Aconnection
may be called a four-wire E&M circuit even though it actually has six toeight physical wires.
The equipment and the type ofservice from the CO determine the type ofaccess signaling on
FXS and FXO ports. Standard home telephone lines use the more common loop-start, but
business telephones can use ground-start lines instead. Both signaling types differ involtage
levels and the grounding of the tipand ring lines. Ground-start signaling reduces glare, a
condition inwhichbothends attemptto seizea trunk at the sametime. Therefore, it is better
suited for lines between PBXs and in businesses with substantial call volume.
The E&M supervisory signaling isclassified as type I, II, III, IV, V, or Signaling System Direct
Current No. 5 (SSDC5). Thesetypes differintheconnections of E&M leads to battery and
ground. E&M interfaces can indicate on- oroff-hook status and telephone line seizure by using
any of the following three types of access signaling:
Immediate-start: Immediate-start is the simplestmethodof F.&M access signaling. The
callingside seizesthe line by goingoff-hookon its E-lead and sends dual tone
multifrequency (DTMF) digits, following a short, fixed-length pause.
Introduction lo Voice Gateways
Wink-start: Wink-start is the most commonly used method for E&M access signaling, and
is the default for E&M voice ports. Wink-start minimizes glare. The calling side seizes Inc
line by going off-hook onits E-lcad. then waits for a short temporary off-hook pulse, or
"w ink." from the other end onits M-lead before sending address information. The switch
interprets the pulse as an indication lo proceed and then sends the D'I'MF digits.
Delay-dial: In delay-dial signaling. Ihe calling station seizes the line by going off-hook on
its F-lead. Altera timed interval, the callingside looksat the statusof the calledside. If the
called side ison-hook. the calling side starts sending information as DTMF digits. Ifthe
called sideis off-hook, thecalling sidewaits until thecalled sidegoeson-hook andthen
starts sending address information.
This subtopic covers analog address signaling.
697 >
770 >
852
941 >
1209 1336 1447
I I I
v *
nil
mmm
Frequency Tone Matrix (Hz]
The dialing phase allows the subscriber to enter aphone number (address) ofatelephone at
another location, fhe customer enters this number witha touch-tone phone that generates tones.
Telephones use two different types ofaddress signaling to notify the telephone company where
a subscriber is catling:
Legacypulse dialing
DTMF dialing
These pulses or tones arc transmitted to the CO switch across atwo-wire twisted-pair cable (tip
and ring lines).
Current analog circuits use DTMF tones toindicate the destination address. DTMF assigns a
specific frequency (consisting oftwo separate tones) to each key on the touch-tone telephone
dial pad. The combination ofthese two tones notifies the receiving subscriber ofthe digits
dialed.
The table inthe figure shows the frequency tones that arc generated by DTMF dialing.
i 2010 Cisco Systems, Inc
Introduction to Voice Gateways J-87
Call Progress Tones
fhis subtopic covers informational signaling.
Call Progress Tones
North American call progress tones example:
Frequency (Hz) |Ori
Continuous Continuous
0.5 0.5
2 4
1 3
0.2 0.3
03 0.2
0.1 0.1
Dial 350+440
Busy 480+620
Ringback, normal 440 +480
Ringback, PBX 440 +480
Congestion (toll) 480 +620
Reorder (local) 480 + 620
Receiveroff-hook 1400 + 2060 +
2450 + 2600
No such number 200-400
Continuous, 1- Continuous, 1-Hz
Hz frequency frequency
modulation modulation
1he FXS port prov ides informational signaling using call progress tones. These call progress
tones are audibleand are usedby the FXS-connected dev ice lo indicate the statusof calls. The
progress tones listed in die table are for North American phone systems. The phone systems in
other countries use adifferent set ofprogress tones. The call progress tones are as follows:
Dial tone: Indicates that the telephone company is ready to receive digits from the user
telephone
Busy tone: Indicates that a call cannot becompleted because the telephone at the remote
end is alreadv in use
Ringback (normal or PBX) tone: Indicates that the telephone company isattempting to
complete a call on behalf of a subscriber
Congestion progress tone: Used between switches toindicate that congestion in the long
distance telephone network currently prevents a telephone call from being progressed
Reorder tone: Indicates that all the local telephone circuits are busy, and thus prevents a
telephone call frombeing processed
Receiveroff-hook tone: The loud ringing that indicates thai the receiver of a phone is left
off-hook for an extended period of time
No such number tone: Indicates that the number dialed cannot be found in the routing
table of a switch
-88 Implementing CiscoVoice Communications and QoS (CVOICE) v8.0
Configuring Analog Voice Ports
This topic describes the configuration ofanalog voice ports on Cisco Unified Communications
gateways.
Configuration Overview
Configuration tasks:
FXS
- CP tones, signaling{loop-start, ground-start)
FXO, FXS-DID
- Signaling (loop-start, ground-start)
E&M (Ear and Mouth)
- Operation (two-wire, four-wire)
Type (1,2.3,5)
- Signaling (immediate-start, wink-start, delay-start)
Centralized Automated Message Accounting
- Signaling (CAMA)
- ANI mapping
There are similarities and differences in the configuration of analog ports.
"Ihe call progress tones and ringcadences areconfigured ontheFXS ports because thephones
are attached to them.
The FXS. FXO. andFXS-DID ports share thesame signaling options: loop-start andground-
start.
The E&M voice ports define theoperation mode (two-orfour-wire), thetype (I, 2. 3. 5),and
the signaling method(immediate-start, wink-start, delay-start)
CAMA trunks areeonfigured for CAMA signaling andwith an ANI mapping definition.
Configuring FXS Voice Ports
This subtopic explains theconfiguration of FXS ports.
Configuring FXS Voice Ports
I
/
^
voice-port 0/2/0
signal loopaCart
cpCone GB
ring cadence patternOl
no ahutdown
The figure shows how 10 configure a voicegalewav to routecalls to a traditional telephone
connected to an FXS port in Great Britain. The port is configured withthesesettings:
FXS voice port position: module II. voice interface card (VIC) slot 2. andvoice port 0.
Supervisor} signaling: loop-start
Call progress tones: specific lo Great Britain
Ring cadenee: pattern 1
The pattern.V.Vkev wordprovides preset ringcadeneepatterns for use on any platform, fhe
define kevword allows vouto createa custom ringcadence. On the router, onlyone or two
pairs of digits can be entered under the define keyword.
Finally. thevoice port canbedisabled using theshutdowncommand or activated using the no
shutdown command.
2010Cisco Systems. Inc
Configuring FXO Voice Ports
This subtopic explains how to configure an FXO port for outbound and inbound private line,
automatic ringdown (PLAR) connections.
Configuring FXO Voice Ports
FXO emulates end-user equipment connectedtothe COswitch.
DIDnot supported on FXO
DNIS not received from CO switch
Incoming callsdirectedto internal numberusingprivateline, automatic
ringdown (PLAR)
voics-port 0/0/0
signal loopatart
connection plar opx 4001
dial-peer voice SO pots
deatination-pattera 9T
port 0/O/0
Iftheoffice has only FXO trunks, thedirect inward dialing capability isnot available. 'Ihe
DialedNumberIdentification Service(DNIS) is signaled for outbound calls only, while
inbound calls carry only theANI. This scenario iscommon fora small standalone office or a
small branch of a bigger network that hasonly a few business tines from the local CO.
Because DID is not supported on FXO trunks, thecallers musteitherbe presented a secondary-
dial tone, or the inbound calls must be autotcrminated on a predetermined destination, most
often the autoattendantor the receptionist extension. This can be achieved withan optional
PI .AR configuration, using the connection plaropx command onthe voice port eonnecting to
the PSTN. The particular destination extension isassociated with thetrunk, and all calls
arriv ing onthat trunk areswitched as if they had dialed theconfigured extension. In this
example, all inbound callsareforwarded tothereceptionist extension 4001.
Configuring FXS-DID Voice Ports
This subtopic describes the configuration of FXS-DID ports.
Configuring FXS-DiD Voice Ports
501-1001 ^/
501-1002
Vo FXQûtoQ^^tE^
voice-port 0/0/0
fligna1 did wink-start
voice-port 0/1/0
signal grounds tart
dial-peer voice 1 pota
direct-iu-atd-dial
destination-pattern S|2-81
port 0/1/0
I'v picallv. FXS ports connedtoanalog phones, butsome carriers otTer FXS trunks that support
DID. An FXS DID trunk can only receive inbound calls, thus a combination of FXS DID and
FXO ports is required for inbound andoutbound calls. FXS-DID supports the same supervisory
signaling asregular FXS and FXO ports: loop-start and ground-start, with ground-start being
the preferred method.
Thediagram inthe figure showsan analogtrunk usingan FXSDID trunk for inbound calls and
a standard FXO trunk for outbound calls.
1-92 Implementing Cisco Voice Communicationsand QoS (CVOICE)
Configuring E&M Voice Ports
This topic explains the configuration of E&M voice ports.
Configuring E&M Voice Ports
/L
\
Irfcoinfl DNIS
Outbouno DNIS
voice-port 1/1/1
aignal nink-atart
operation 2-wlre
type 1
do shutdown
exit
iacoming called-number
direct-imcard-dial
destination-pattern 1...
forward-digito all
port 1/1/1
The key configuration parameters of anE&M analog trunk areas follows:
The F&M signaling type
Two- or four-wire operation
The F&M type
The figure shows anexample of anE&M portthat provides connectivity toa PBX. In this
example, the voice gateway connects via an E&M trunk toa PBX toallow the IP phones tocall
the POTSphones using a four-digit extension.
"fhe voice port is configuredwith these settings:
Signaling: wink-start
Operation mode: two-wire
E&M signaling: Type I
Both sides of the trunk need to have a matching configuration, fhis example configuration
showsan F&Mtrunkusingwink-start signaling, E&M Type I, and two-wire operation.
) 2010 Cisco Systems, Inc
Configuring CAMA Voice Ports
This subtopic describes CAMA trunkconfiguration.
Configuring CAMA Voice Ports
m
4.
T1 PRI for Staridaracalis

PSTN
for Emetgercy
Calls
fii
Public Safely
Answering
Point
vw;-port 1/1 '1
ml napping 1 312
signal ciu U-KP$-iax-im -ST
dl*l-p*BJ: vgIcb 1 pots
lncofoiog oal led- nuntor
direct-inward-dil
dial-pear volna 50 poto
daiciuius.pituiii S,'J-S]
port 0/0/D.23
dial-p*r voico 511 pota
dnstlnation-pattarn 911
poet 1/1/1
rorwsrd-dlgita .11
dul-piar vole. 9911 pota
d.ltlUltl-ptt.rn 9511
port 1/1/1
forward-dlglto 3
Centralized Automated Message Accounting(CAMA) available in North America only
The figure shows an enterprise with a 11 PRI circuit for normal inbound and outbound PSTN
calls. Because the local PSAPrequires a dedicated CAMAtrunkfor emergency (911) calls, ail
emergency calls are routed using a dial peer pointing to the CAMA trunk.
The voiceport !/[/] is the CAMA trunk. Theactual configuration depends on ihe PSAP
requirements. In this case, the digit 1is usedto signal the area code 312. The voiceport is then
configured for CAMA signaling usingthe signal cama command. I'our optionsexist:
KP-O-NXX-XXXX-ST: 7-digit ANI transmission. Ihe Numbering Plan Area (NPA) or
area code is implied by the trunk group and is not transmitted.
KP-O-NPA-NXX-XXXX-ST: 10-digil transmission. 1he E. 164 number is fully
transmitted.
KP-O-NPA-NXX-XXXX-ST-KP-YYY-YYY-YYYY-ST: Supports CAMA signaling with
ANI and Pseudo ANI (PANI).
KP-2-ST: Default transmission whenthe CAMAtrunkcannotgel a corresponding
Numbering Plan Digit (NPD) in the lookuptable, or when the calling number is fewer than
in digits. {NPA digits are not available.)
KP-NPD-NXX-XXXX-ST: 8-digit ANI transmission, where the NPDis a single
multifrequency (MK) digit that is expanded into the NPA. The NPD table is
preprogrammedin the sending and receiving equipment.
Inthisexample, the PSAPexpects NPDsignaling, withthearea code312 beingrepresented by
the digit 1. CAMA is available in North America only.
Digital Voice Ports
fhis topic describes the types ofdigital voice ports and their characteristics.
Digital Voice Ports Overview
T1: Uses time-divisionmultiplexing (TDM) to transmit digital
data over 24 voice channels using channel associated
signaling (CAS)
E1: Uses TDM to transmit digital data over 32 timeslots,
including 30 voicechannels, 1 framingchannel, and 1
signaling channel
ISDN: Acircuit-switched telephone network system designed
to allow digital transmission of voice and data over ordinary
telephone copper wires:
BRI: 128 kb/s; 2 B channels and 1 D channel
- T1 PRI: 1.472 Mb/s; 23 B channelsand 1 Dchannel
E1 PRI: 1.920 Mb/s; 30 B channels and 1 D channel
- Uses common channel signaling (CCS)
Digital voice ports arefound at theintersection of apacket voice network and a digital, circuit-
switchedtelephone network. The digital voiceport interfaces that connectthe router or access
server to "11 or E1 linespass voicedata and signalingbetweenthe packet networkand the
circuit-switched network.
There are threetypes of digital voicecircuitsthat are supported on Ciscovoicegateways:
Tl: Usestime-division multiplexing (TDM)to transmitdigital data over 24 voicechannels
usingchannel associated signaling (CAS) or 23 voice channels usingcommon channel
signaling (CCS)
Fl: Uses TDM to transmit digital data over 30 voice channels using either CAS or CCS
ISDN: A circuit-switchedtelephone network system using CCS
BRI: 2 bearer (B) channels and 1 data (D) channel
Tl PRI: 23 B channels and I D channel
El PRI: 30 B channels and 1 Dchannel
Digital Circuit Types
This subtopic prov ides a classification of Ihe common digital circuit types.
Digital Circuit Types
ijMjm|
Circuit OpLon ISKi&tfi"tl^MtWMBi^|ffi^^ffl|H
Digital l'TICAb
Analog signaling over digrtai T1rTt
Car provide ANI calling party ID(caller ID]
srjh
CCSj
'IPRI
r-PBt
More services than CAS
Separate signaling channel (D channel)
Common on modem PBXs
I'HNFftS Multiple ISDN PRI inlerfaces controlled by a
single D cnannel
Backup D channel can be configured
BV
Moslly for Europe Middle East, and Africa
cso
Created lor mteraperation of PBXs from
different vendors
Rich in supplementary services
The table in the figure lists some of the common digital circuit options.
Tl and FI have existed since the early voice networks. They carry multiple calls across one
copper loop. Tl/Eil circuitsuse TDMto transmit digital data insteadof theold analogsignals.
A Tl orTl ISDN PRI line contains 24 full-duplex channels or lime slots, and an Fl line
contains 30 full-duplex channels. The signal on each channel is transmitted at 64 kb/s. a
standard known as digital service level 0(DS0): the channels are known as DSOchannels. The
dsO-group command creates a logical voiceport (a DSO group) from someor all of the DSO
channels. This logical voice port allows youtoaddress those channels easily, as a group, in
voice port configuration commands.
Ihe method that is used to transmit the informationdescribes the way in which the emulated
analog signaling is transmitted over digital lines.
Digital lines use two types of signaling:
CAS: lakes place within the voice channel itself
CCS: Sends signaling information down a dedicated channel
Two main types of digital trunks with CAS exisl:
Tl CAS trunk: 'fhis type of circuit allows analog signaling via a digital Tl circuit. There
are mam CASvariants that operateover analogand digital interfaces. Acommon digital
interface is Tl or hi (European version), where each channel includes a dedicated
signaling element(also called"robbed-bit signaling" on 'I'Is). The type of signaling that is
most commonly used with Tl CAS is E&Msignaling. In addition to setting up and tearing
down calls. CAS prov ides the receipt and capture of DNISand ANI information. The main
disadvantage of CASsignaling is its use of user bandwidth to perform thesesignaling
functions.
1-96 Implementing Cisco Voice Communications and QoS (CVOICE] v8 0 i 2010 Cisco Systems, Inc
El R2 trunk: R2 signaling isaCAS system that was developed in the 1960s and is still in
use today in Europe. Central and South America. Australia, and Asia. R2 signaling exists in
several country versions or variants in an international version called ITU-T R2. The R2
signaling specifications are contained in ITU-T Recommendations Q.400 through Q.490.
R2 also provides ANI.
>2010CiscoSystems. Inc. Introduction to Voice Gateways 1-97
T1 CAS Overview
This topic describes Tl CAS trunks and associated signaling.
T1 CAS Overview
T1 CAS uses in-band robbed-bit signaling.
s Signaling for a particular traffic circuit is permanently
associated with that circuit.
4 Signaling is based on analog signaling: loop-start, ground-
start, and E&M variants
E&M supports various feature groups.
No D Cnannel Required
Analog Signaling
Tl CAS uses a digital Tl circuit with in-band signaling information. CAS on Tl is
accomplished by usingbits in theactual voicechannel to transmit signaling information. CAS
is sometimes called robbed-bit signaling because the network robs user bandwidth for
signaling. A bit is taken from every sixth frame of voice data to communicate on- or off-hook
status, wink-start, ground-start, dialed digits, and other information about the call.
TI CASusesthe samesignaling ty pesthat are av ailable for analogtrunks: loop-start, ground-
start, and E&M variants such as wink-start and immediate-start. There are also various feature
groups that are available whenyou use E&M. Hereare some common feature groups:
E&M Feature Group (1 (FGB): Provides inbound and outbound DNIS, and inbound ANI
{only on Cisco AS5300 Series Universal Access Servers. Cisco AS5400 Series Universal
Gateways, and Cisco AS5800 Series Universal Gateways)
E&M Eeature Group D (EGI)): Provides inbound and outbound DNIS. and inbound ANI
E&M FGD-Exchange Access North American (EANA): Provides inbound and outbound
DNIS. and outbound ANI
Implemenling Cisco Voice Communications and QoS (CVOICE) v8 0
)20I0 Cisco Systems. Inc.
T1 CAS SF Format
This subtopic explains the TI CAS Super Frame (SF) format.
T1 CAS Super Frame Format
I TimeSlot
8 Bits
[ 24 Bbits * 1bit =1frame (193 bits)
l|g|3J4Ja|6J7J8}9 110 111 jl2|l3|l4Jls|l6|f7[lB|iejao|2l|22)z3J24
1 Bit Syr
| 12 Frames - Super Frame |
1 Z 3 4 8 7 8 9 10 11
24-]7 bits* I robbed bit) +1bit =1frame(193 bits) |
I I I I I I I I I I I I I I I I I I I I I I I II
Time Slol
7 Bits'
1 RoODed Bit
The figure shows the CAS with the Tl SF format. The top row ofboxes represents asingle Tl
frame with 24time slots of 8 bits each. An additional bit, usedto synchronize the SF. is added
at theendof eachframe. Asequence of 12Tl frames makes uponeSF. CAS is implemented
by robbed-bit signaling inframes 6and 12 inthis sequence. Thebottom row of boxes
represents TI frames 6and 12. The least significant bitofeach voice channel isrobbed, leaving
7 bits for voice data.
T1 CAS ESF Format
This subtopic explains the Tl CAS Extended Super Frame (ESE) formal.
| 24'8 bits* 1brt =1frame (193 bits) j
1 2 3 I 5 6 7 8 a to | 11 112 J13 14 15 18 17 18 19 20 21 22 23 24
| 24 Frames =Extended Super Frame
| 9 i0l11BF^ H15 16 |17 H119 |20 f21 |22 |23
12 3 4 5 WM 7 8
24 * (7 bits +- 1 robbed toil) + 1 but = 1 frame (193 bits)
1 Bit
Sync
1111111 11111111 n n
Time Slot
7 Bits -
1 Robbed Bit
"fhe FSFformat wasdeveloped as an upgrade to SFandis now dominant inpublic andprivate
networks. Bothformats retainthe basicframe structure of one framing bit followed by 192 data
bits. However. FSF repurposes the use of the F bit. In ESE. of the total 8000 F bits that are used
in'['1 within a second (8-kHzsampling). 2000are used for framing, 2000are usedfor cyclic
redundancv check(CRC) for error checking onlv. and 4000 are usedas an intelligent
superv isorv channel to control functions end-to-end (suchas loopback and error reporting).
1-100 Implemenling Cisco Voice Communications and QoS (CVOICE) v8 0 2010 Cisco Systems, Inc
E1 CAS Overview
Thissubtopic explains theFT CAS signaling.
E1 CAS Overview
E1 uses multiframe format:
Consists of 16 consecutive 256-bit frames
Consists of 32 time slots:
- One slot for frame synchronization
- One slot for signaling
- 30 slots for actual voice traffic
E1 R2 variant supports inbound and outbound DNIS and ANI
- Other variants may not support inbound ANI
No D Channel
Analog Signaling
30 BGhannels (Voice)
^
\
An E1 circuit is like a Tl circuit. It is a TDM circuit that carries several DSOs in one
connection. El circuits arewidely used in Europe, Asia, and Central and South America.
One large difference between an El and aTl is that an El bundles 32 time slots instead of24.
This difference results in a bandwidth of 2.048 Mb/s for an El. With an El circuit, one time
slotis used for framing and oneis used for signaling, leaving 30time slots available for user
data.
You may use different types ofsignaling over an El interface, either CAS orCCS. One ofthe
common CAS signaling variants is El R2. El R2 signaling supports inbound and outbound
DNIS and ANI. El R2 uses the El multiframe format.
E1 CAS Multiframe Format
This subtopic explains the El CAS multiframe format.
E1 CAS Multiframe Format
Muflrframe
2 046 Mb.-s
il.>MM
1 Frame Start
I Frame Signaling for
! Frame Srqnahnqfor
I France Siqnaimqto
. Frame Signaling to
: Frame SiqnalinotoT
Ffame Sioialuiq for
i Frame Signaling for
i J-ra^w Siqnalmq for
Frame Siqnaf.nqtor
Signaling tor
Signaling to'
Siqnahiq tor
Siqnalnq for
Siqnahna tor
Signing for
TT
2
3
4
t
6
7
e
9
10
11
2
13
14
15
16
i1iMj.M||m|m !, 28 ?9 ill 11 32
lolMultiframe 1 j
r \*FCeStots2 andlB
r \*>ice Slo1s3 and19
r\*iceSlols4 and20
r WnceSlots5anfl21
WiceSlotB6an<!22
Wjree Slots? and 23
WiceSlotsB and?4
*iceSlots9 and 25
WiiceSlotslO and26
Mjice Slots 11 and 27
*iceSi(it5l2and28
*iceSlots13 ana 29
*iceSlo1s14 and 30
VaiceSlolsl5 and 31
W3iceSlots16 and 33
A multiframe consists of 16 consecutive 256-bit frames. Each frame carries 32 time slots. The
first time slot is used exclusive!) for frame synchronization. Time slots 2 to 16and 18lo 32
carrv the actual voice traffic, andtimeslot 17is used for signaling.
Note There is often confusion about the signalingtimeslot position. Timeslot 17 means that the
first timeslots started to be counted as first, while somelimes it has the position 0, which
would lead lo the signaling lime slot position 16.
The first frame in an El multiframe includes the multiframe format information in time slot 17.
Frames 2 to 16 include the signaling information, each frame controlling two voice time slots.
The figure shows die signaling concept that is used by El. limeslot 17 is used for signaling,
and each of its frames carries infomialion for two voice time slots. This results inthe following
frame allocation for signaling:
1. Krame, time slot 17: Declares the multiframe
N. Krame, time slot 17: Signaling for time slots N and N+16. where N is from 2 lo 16
1-102 Implementing Cisco Voice Communications and QoS (CVOICE] v8.0
Understanding ISDN
Thistopic describes the ISDN signaling, interface, andfeatures.
ISDN
Common Channel Signaling (CCS)
Voice, video, and data are sent over separate channels.
. '~?~&Ptt~~ DChannel 16 kb/s (Signaling)
2 B channels fi/oice)
D Channel 64 kb/s (Signaling)
23 B Channels (Voice)
D Channel 64 kbfe (Signaling)
30B Channels (Voice)
Anotherprotocol that is usedfor digital trunks is ISDN. ISDN is a circuit-switched telephone
network systemthat is designed to allowdigital transmission of voice and data over ordinary
telephone copper wires. This design results in better quality andhigher speeds than is available
with the PSTN system.
ISDN comprises digital telephony anddata-transport services offered by regional telephone
carriers, ISDNinvolves the digitizationof the telephone network, which permits voice, data,
text, graphics, music, video, and other sourcematerial to be transmitted over existingtelephone
wires.
ISDN Services
In contrastto CASsignaling, which providesonlylimitedfunctionality, ISDN oilers additional
supplementary services such as call waiting andDoNot Disturb (DND). ISDN applications
include high-speed image applications (such as Group 4 [G4] fax), additional telephone lines in
homes to serve the telecommuting industry, high-speed file transfer, and video conferencing.
Voice service is also an application for ISDN.
ISDN Media Types
Ciscoroutingdevices support ISDN BRI and ISDN PRI. Bothmediatypes use Bchannels and
Dchannels. The B channels carry user data. The Dchannel provides an out-of-band signaling
channelthat carriessignalinginfonnation for all Bchannels. Therefore, the signalingtype used
by ISDN is CCS.
ISDN BRI: (Referred to as "2 B + D")
Two 64-kb/s Bchannels that carry voice or data for a maximumtransmission speed
of 128 kb/s
One 16-kb/s Dchannel thatcarries signaling traffic, that is. instructions about how-
toprocess each of the Bchannels, although it cansupport userdatatransmission
under certain circumstances
The Dchannel signaling protocol comprises I.avers I through 3 of theOpen Systems
Interconnection (OSI) reference model. BRI alsoprov ides for framing control andother
overhead, bringing its total bit rate to 192kb/s.
BRI is very common inEurope and isalso available in North America. BRI allows up to
two simultaneous calls.
ISDN PRI: (Referred to as "23 B t D" or"30 B - D")
23 Bchannels (in North America and Japan) or 30 Bchannels (in the rest of the
world) that earn voice or data >ieldinga total bit rate of 1.544 and 2.048 Mb/s.
respect iveh
One 64-kb/s Dchannel that carries signaling traffic
Note PRI is economically preferable to BRI because there is usually an interface cardsupporting
PRI already in place on modern PBXs
ISDN BRI and PRI Interfaces
This subtopic describes theISDN BRI and PRI interfaces.
ISDN BR! and PRI Interfaces
6 Channels
D Channels
Framing Rales
Frame Formats
Line Coding
Country
2x64 kb/s
1 x 16 kb/s
16 kb/s
NT, TE frame
23 x 64 kb/s
1 x 64 kb/s
8 kb/s
SF, ESF
2B1Qor4B3T AHIorBSZS
World North America, Japan
30x64 kb/s
1x64 kb/s
64 kb/s
Multiframe (CRC-4)
HD83
Europe, Australia
The figure illustrates thedifferent capabilities of BRI andPRI interfaces.
Using ISDN for voicetraffichas these benefits:
ISDN is perfect for G.711 pulse code modulation (PCM) because each Bchannel isa
complete 64 kb/s with no robbed bits.
ISDN has a built-in call control protocol known as ITU-TRecommendation Q.931.
ISDN can convey standards-based voice features, such as speed dialing, automated
operator services, call waiting, call forwarding, and geographic analysis of customer
databases,
ISDN supports standards-based enhanced dial-up capabilities, such asG4fax and audio
channels.
Line Encoding
Digital Tl/El interfaces require that line encoding is configured tomatch thatof the PBX or
CO that is being connected to the voice port. Line encoding defines thetype of framing that is
used on the line.
BRI lineencoding methods include two-binary, one-quaternary (2B1Q) and four-binary, three-
ternary (4B3T). 2B IQis a physical layer encoding used for ISDN BRI. 2B1Q uses four signal
levels, which are -450. -150, 150. and 450 mV, each "IQ" equivalent to two binary bits (2B).
4B3Trepresents four binary bits usingthreepulses.
Tl PRI line encoding methods include alternate mark inversion (AMI) and binary 8-zero
substitution (B87S). AMI is used onolder Tl circuits and references signal transitions with a
binary I. or -mark," B8ZS. a more reliable method, is more popular and isalso recommended
for PRI configurations, B8ZS encodes asequence ofeight zeros in a unique binary sequence to
detect line-coding violations.
Supported 1-1 lineencoding methods are AMI andhigh-density bipolar 3 (IIDB3). which is a
form of zero-suppression line encoding.
1-106 Implementing Cisco Voice Communications and QaS (CVOICE) v8.0 2010 Cisco Systems, Inc.
ISDN Architecture
This subtopic explains theISDN protocol layers and theirfeatures.
ISDN Architecture
Protocol
name
Features
Task
Layer 2
Q.921
Link Access Procedure on the D
channel (LAPD). Simlarto HDLC.
Provides error detection and
correction.
Provide terminal endpoint identifiers
(TEls) as Layer 2 addresses to end
devices:
through static configuration
- dynamical* allocated by PSTN
0.931
Various message types:
- cal-eslabtbhment
- cal-termi nation
information
- miscelaneous
Supports these connections:
user-to-user
- circuit-swSched
packet-switched
Layer 2 of theISDN signaling protocol is Link Access Procedure on the Dchannel (LAPD),
LAPD is similarto High-Level DataLinkControl (IIDLC) and Link Access Procedure,
Balanced (LAPB). Themost important feature of LAPD is its capability todetect andcorrect
errors. As theexpansion of the LAPD acronym indicates, this layeris used across the D
channel to ensurethat control and signalinginformation flows and is received properly. The
LAPD frame format is similar to that of HDLC. Like HDLC, LAPD uses supervisor,'
infonnation and unnumbered frames. The LAPDprotocol is formally specified in ITU-T
Recommendation Q.920and ITU-T Recommendation Q.92I. TheTerminal Endpoint Identifier
(TEI) fieldidentifieseither a singleterminalor multipleterminals. ATEI of all Is indicates a
broadcast.
TwoLayer 3 specifications are used for ISDN signaling: ITU-T Recommendation 1.450 (also
known as Q.930) and ITU-T Recommendation 1.451 (also known as Q.931). Together, these
protocols support user-to-user, circuit-switched (the Bchannels), and packet-switched (the D
channel)connections. Avarietyof call-establishment, call-termination, information, and
miscellaneous messages are specified, including SETUP, CONNECT. RELEASE. USER
INFORMATION. CANCEL. STATUS, and DISCONNECT.
Non-Facility Associated Signaling
This topic describes dieISDN Non-facility Associated Signaling (NEAS) feature.
Allows a single Dchannel to control multiple PRI interfaces.
Abackup D channel can be configured, but only the NFAS
primary D channel must be configured,
NFAS is only supported with a channelized T1 controller.
iSDN Tl PRI .SFAS
23 B Channels (Voce)
23 BChannels Choice) -'. ,
24 e Channels (Voice) V
ISDNNFAS allows a single Dchannel to control multiple PR! interfaces. Use of a single D
channel to control multiple PRI interfaces frees one Bchannel on each interface to cany other
traffic. A backup I) channel can be configured for use when the primary NFAS Dchannel fails.
When a backup D channel is configured, any hard system failure causes a switchover to the
backup I) channel and currently connected calls remain connected.
NFAS is supported onlv with a channelized Tl controller and. as a result, must be ISDN PRI-
capablc. After the channelized 11 controllers arc configured for ISDN PRI. only the NEAS
priman Dchannel must be configured: its configuration is distributed to all members of the
associated NEASgroup. Any configurationchanges made to the primary Dchannel will be
propagatedto all NFAS group members. Fheprimary Dchannel interface is the only interlace
shown after ihe configuration is written to memory.
Ihe channelizedTl controllers must be configured for ISDN. The router must connect to either
an AT&T4ESS. Nortel DMS-100 or DMS-250. or National ISDNswitch type.
The ISDN switch must be provisioned for NFAS. The primary and backup Dchannels should
be configured on separate II controllers. The primary, backup, and B-channel members on the
respective controllers should be the same configurationas that configured on the router and
ISDN switch. The interface ID assigned to the controllers must match the interface ID of the
ISDN switch.
You can disable a specified channel or an entire PRI interface, thereby taking it out of service
or placing it intoone of the other states that are passed lo the switch, using the isdn sen ice
interface configuration command.
Implementing Cisco Voice Communications and QoS (CVOICE] v8 0 2010 Cisco Systems, Inc.
Configuring Digital Voice Ports
Thistopic describes howto configure thedigital voice ports.
Configuring Digital Voice Ports
T1/E1 Voice Port Configuration Overview
Configure controller settings:
Framing
Line encoding
Clock source
Create digital voice ports:
DSO group
- Time slots
Signal type
Configure voice port parameters:
compand-type
cptone
Whenconfiguring a Tl or El trunk, you must definethesekey parameters:
Framing formats
Fine encoding
Clock source
DSOgroups
Uponsuchconfiguration, a logical voiceport is created.That voiceport maybe fine-tuned with
parameters such as compand-type or cptone.
Framing Formats
The framing format parameter describes the waythat bits are robbedfrom specific frames to be
used for signaling purposes. The controller must be configured to use the same framing format
as the line from the PBX or CO that connects to the voice port that you are configuring.
El lines can be configured for cyclic redundancy check 4 (CRC4) or no CRC. with an optional
argument for El lines in Australia.
Line Encoding
Digital Tl/El interfaces require that line encodingis configured to matchthat of the PBXor
COthat is beingconnected to the voiceport. Lineencodingdefines the type of framing that is
used on the line.
Configuring Digital Ports
Thissubtopic explains how to configure voiceportsconsisting of digital voiceslots.
Configuring Digital Ports
DSOGroup Configuration
Network Module Slot 1
VWIC Slot 0
ConfiguresTl
Controller 1/0
controller tl 1/0
framing eaf
clock source line
linacode bBis _
dsO-group 1 t
T1
Creates DSO
Group, or Logical
Voice Port, 1/0 1
by Grouping 12
Time Slots
Together
lots 1-12 type eSra-wink-start
You must create a digital voice port on the Tl/El controller to be able to configure voice port
parameters. You must also assign time slots and signaling to the logical voice port through
configuration. The first step is to create the Tl or El digital voice port with the dsO-group dsO-
group-no timeslots tinwslot-Ust type signal-type command, fhis voice is commonly referred to
as a logical \ oicc port, because it encompasses a selected number of voice channels that are
phvsicallv transported over the circuit.
Note The dsO-group command automatically creates a logical voice port that is numbered as
slot/port.dsO-group-no
'fhe dsO-group-no argument identifies the DSO group (numbered from ((to 23 for'!'1 and from
0 to 30 for El). This group number is used as part of the logical voice port numberingscheme.
fhe timeslots command allows \ou to specifv which time slots will belong to the DSO group.
The timeslot-list argument is a single time slot number, a single range of numbers, or multiple
ranges of numbers separated bv commas.
Fhe type command defines the emulated analog signaling method thai llic router uses to
connect to the PBX or PSTN. The type depends on whether ihe interface is Tl or El.
To delete a DSO group, vou must first shut down the logical voice port. When the port is in
shutdown stale, you can remove the DSO group from the Tl or El controller with the no dsO-
group dsO-group-no command.
The figure shows how a dsO-group command gathers some of the DSO lime slots from a TI
line into a group that becomes a single logical voice port, which can later be addressed as a
single entity in \ oicc port configurations. Other DSO groups for voice can be created from ihe
remaininglime slots shown in the figure, or the time slots can be used for data or serial pass-
throLisdi.
1-110 Implementing Cisco Voice Communications and QoS (CVOICE) v8.0 2010 Cisco Systems, Inc.
Configuring Digital Ports (Cont.
Clock Sources
;ootroller el 1/0
training erct
liDecode hdb3
cloeH source intern.l
dsO-group timeBlotB 1-15 type a
controller tl 1/0
framing b
linecode ami
clock aourca line
tH-uink-8trt
Clock
V1
PBX
Clock Sources
eonfiTured can be external (from the line) or internal to the router digital interface.
If the timing source is internal, timing derives from the onboard phase lock loop (PLE) chit. m
dig 1vîce interface. If the timing source is line (external), then timing ^^
PBX oPSTN CO to which the voice port is connected. It is generally preferable to derive
Eminem*e PSTN because its clocks are maintained at an extremely accurate level.
ES^mUis the default setting for the clocks. Wheri two ot^more conh,ers>
configured, one should be designated as the pnmary clock source, .1 will d.rect ihe other
controllers.
The figure depicts the following types of clocking:
. Single voice port providing clocking: In this scenario, the router internal oscillator is the
clock source The voice card is synchronized to the internal oscillator and dire* the
clocking on the line. ACisco VoIP gateway rarely provides clock.ng to the CO because CO
clocking ismuch more reliable.
. Sing.e voice port receiving interna. Cocking: In this scenario the digM voice: hardware
receives clocking from the connected device (CO telephony switch or PBX). Ihe 11 L
clocking'sdirected by the clock reference on the receive (Rx) side of the digital fine
connection.
12010 Cisco Sysiems, Inc
Introduction lo Voice Gateways
1-111
1-112
is*;r:"and recmc siena'bccomes sn *"-,hc"e^::;rsPin
To elim,ale the clocking mismatch, change ,hc default clocking behavior through Cisco IOS
tISlTn'a*T,K ".""*-'^*IP.. command am The ,er
touse mt clock tram the fine via the spec edslot WAN interface r-irH ru/irv , i V
n,cgrat,on module (A.M, and s.vnchroLe the onboard dock ^êêferen^ ^
netnork-clock-participate |slo slo, number Iwic nic-sht | aim aim-slot-number}
nctork-cIock-select/*w/n |bri | tl | el} slolipor, *
l^l^V'WSr 'T^^ inSti'"Cd-lllC Cmm^ mu b< "P*** ^ *
c^'mand ' "g ^ ^ CnnmiCd US"18 the show network clocks
Implemenling Cisco Voice Communications and QoS (CVOICE) v8 0 ,,.. Z
' (9^010 Cisco Systems, Inc
Configuring Digital Ports (Contj
Logical Voice Port Configuration
voice-port 1/0:1
cptone US
compand-type u-law
no shutdown
T1 CAS
&M Wink-Start
PSTN
Logical Voice Port Configuration
The figure shows howto configure the logical voice port that is created using the DSO group on
a Tl/El controller.
After setting up the controller, you can now configure voice port parameters for that digital
voice port. When you specified a dsO-group, the system automatically created a logical voice
port. You must then enter the voice-port configuration mode to configure port-specific
parameters. Each voice port corresponds to one dsO-group command.
To configure basic parameters for digital voice ports, enter voice-port configuration mode using
the voice-port s\otJport:dsO-group-number command. You can fine-tune options, such the
voice call progress tones and other locale-specific parameters to be used on this voice port, and
the companding standard that is used to convert between analog and digital signals. The
compand-type options are mu-law or a-law.
Note The compand-type command is used Incases when the DSP Is not used, such as local
cross-connects, and overwrites the compand-type value set by the cptone command.
Configuring Digitai Ports (Cont)
T1 CAS: Inbound E&M FGD and Outbound FGD-EANA
l~\~U
E&MFGD" Time Slots T to 12, Receive ANI
PSTN
H
E5MFGD EANA" Time Slots 13(0 24, Send ANI
controller Tl O/O/O
framing eat
linecode b6ie
dsO-group D times-lots-
dsQ-group 1 timeslots
incoming called -numbe
direct-inward-dial
dial-peer voice 90 pot
destination-patt
port 0/0/0: 1
1-12 type asn-fgd
13-24 type fgd-ean
ST
"FGD = E&M Feature Group D provides inbound and outbound DNIS. and inBoundANI
"EANA = FGD-Excftange Access Norm Amencan. provides inbound and oulbound DNIS,
and njtbmnd ANI
T1 CAS: Inbound E&M FGD and Outbound FGD-EANA
Fhe figure provides a configuration example for two DSO-groupsdefined on a single Tl
controller, one used for outbound and one for inbound calls.
Because E&M FCiD supports onlv inbound AM. a deployment requiring both inbound and
outbound AM can combine an E&M FCiDand FGD-EANA trunk. The figure provides a
configuration, where the EGD trunk is used for inbound calls, and the FCiD-EANAtrunk is
used for outbound calls.
1-114 Implementing Cisco Voice Communications and QoS (CVOICE) v8.0 ) 2010 Cisco Systems. Inc.
E1 R2 CAS
Configuring Digital Ports (Cont]
E1 R2 CAS
E1 R2 CAS
Inbol
Outbound CWte'Snd ANI
PSTN
--
v
controller el 0/0/0
dsO-group 0 timeslots 1-31 type r2-digital r2-compelled aai
Incoming called-number ,
direct-invard-dlal
destination-pattern 9T
port 0/0/0:0
The figure provides an example ofhow toconfigure a logical digital voice port using El R2
CAS signaling onan El controller. The dsO-group controller command isconfigured tocreate
the DSO group with the defined signaling type. The Cisco implementation ofR2 signaling has
DNIS support enabled by default. Ifyou enable the ANI option, DNIS information isstill
collected. Specification of the ANI option docs not disable the DNIS collection.
Configuring ISDN
This topic describes how toconfigure ISDN PR] and BRI interfaces, and PRI QSIG signaling.
Configuring ISDN
BRI:
Interface configuration:
- isdn switch-type
isdn incoming-voice
* isdn protocol-emulate
- PRI
Gtobal/controller configuration: isdn switch-type
T1/E1 controller configuration: pri-group
D-channel (interface) configuration: isdn incoming-voice
QSIG signaling configuration
Many PBX vendors support either I'I/El PRI or BRI connections.
To configure PRI or BRI interfaces on a voice galewav, you need to define the ISDNswitch
type, the Dchannel settings, and. optionally, specify the ISDN protocol emulation.
When provisioning ISDNcapabilities for Tl or E] PRI, you first need to enter the basic
configuration of theconlrollers. After theclock source, framing, andlinecodearedefined, you
need to configure the ISDNswitch type and PRI group using these commands:
isdn switch-type: Defines the type of the ISDN switch that the gateway is connected to.
The settingcan be appliedin global, controller, or interface configuration mode. The
settingconfigured in a morespecificmodeoverridesthe moregenericsetting. The
supported types are listed in the table.
Switch Type Description
basic-itr6 German 1TR6 ISDN switches
basic-5ess AT&T basic rate switches
basic-dms100 NT DMS-100 basic rate switches
basic-net3
NET3 ISDN and Euro-ISDN switches (UnitedKingdomand others), also called E-
DSS1 orDSSI
bastc-nti National ISDN-1 switches
basic-nwnet3 Norway Net3 switches
basic-nznet3 New Zealand Net3 switches
basic-ts013 Australian TS013 switches
rone No switch defined
1-116 Implementing Cisco Voice Communications and QoS (CVOICE! v8 0
>2010Cisco Systems. Inc
Switch Type
Description
ntt
Japanese NTT ISDN switches (ISDN BRI only}
primary-4ess
AT&T 4ESSswitch type forthe United States (ISDN PRI only)
primary-5ess
AT&T 5ESS switch typefor the United States (ISDN PRI only)
primary -dmslOO
NT DMS-100 switch typeforthe United States (ISDN PRIonly)
primary-net5
Net5 ISDN PRI switches (Europe)
primary-ntt
INS-Net 1500 for Japan (ISDNPRI only)
primary-ts014
Australian TS014 switches {ISDN PRI only)
vn2
French VN2 ISDN switches (ISDN BRI only)
vn3
vn4
pri-group: Specifies an ISDN PRI group on achannelized Tl orEl controller. The group
consists of thetime slotsthat arcdefined in thiscommand. Onevoice channel is released
for the ISDN PRI signaling purposes.
isdn incoming-voice: Configures the interface for sending all incoming calls tothe DSP
card for processing.
QSIG signaling: Configures the use ofQSIG signaling on the Dchannel. You typically
use this setting when connecting via ISDN toa PBX. The command toenable QSIG isisdn
switch-type primary-qsig for PRI connections and isdn switch-type basic-qsig for BRI
connections.
ISDN BRI Configuration
This subtopic explains the configurationof an ISDN BRI interlace.
Pl-onei-i
2C01
Phcne1-2
2002
PSTN:
neework-clock-parfcicipate wic 0
interface bri 0/0
isdn switch-type basic-net3
isdn overlap-receiving
isdn incoming-voice voice
isdn protocol-emulate user
In this scenario. Routerl is configured witha BRI connection to the PS'FN, providing two B
channels and one Dchannel. These parameters are defined:
The DSP clocking is synchronized with the WIC in slot 0.
The ISDN switch type is set to basic-net3. which is appropriate for most Furopean
countries, fhe isdn switch-type command canbe enteredin global configuration modeor
at the interface level. If you configure both, the interface switchtypetakes precedence over
theglobal sw itch t\ pe. This parameter must match theprovider ISDN switch. Thissetting
is required for both BRI and PRI connections.
The isdn overlap-receiving command allows the incoming numbers to be received digit-
by-digit and not en bloc.
Ihe isdn incoming-voice voice command defines incoming calls as voice-onlv. All
incoming calls are sent to the DSP resources.
The isdn protocol-emulate user command is the default setting and therefore does not
appear in the configuration. Fhiscommand allows the gateway to emulate either the user or
the network side. For two ISDN devices to intcroperate. one must be configured as the
network side, and the other as the user side. Because the COswitches providethe network
side, voice gateways connected tothe ISDN COswitches should keep thedefault setting.
IN
ISDN E1 PRI Configuration
This subtopic explains the configuration ofan ISDN PRI interface.
ISDN E1 PRI Configuration
network-clock-participate wic 0
network-clock-select 1 el 0/0/0
iedn switch-type primary-net5
controller el 0/0/0
pri-group timoalotB 1-31
interface Serial0/0/0:15
isdn overlap-receiving
Inthisscenario. Routerl is configured witha PRI connection tothe PSTN, providing 30
B channels and 1 Dchannel. These parameters are defined:
a fhe DSP clocking is synchronized with theWIC in slot0.
The ISDN switch type isset primary-net5. 'Fhis isappropriate for most European
countries.
The line coding needs tobedefined for the EI controller. In this case, use linecoding ami.
This is not shown in thefigure because this is thedefault configuration.
The framing needs to be defined for the EI controller. In this case, use crc4 framing. This
is not shown inthe figure because this isthedefault configuration.
The clocksourcewill be set to the PSTN. This is the default setting, so it is not shownin
the configuration.
The logical voice ports need tobecreated. This isdone with the pri-group timeslots
command. Youcanconfigure the PRI groupto includeall available time slots, or youcan
configure only a select groupof lime slots.
Thedefault for El is pri-grouptimeslots 1-31. Thisdefines 30 Bchannels as PRI
group time slots (1-15 and 17-31) and allocates time slot 16 asthe Dchannel.
Because the gateway counts the channels starting from 0, the Dchannel isnumbered
15 and the logical voice port name is Serial-Controller ID: 15, for example.
Serial0/0/0:15
The default for Tl is pri-group timeslots 1-23. This defines 23 Bchannels as PR!
group time slots (1-23) and allocates time slot 24asthe Dchannel. Because the
gateway counts the channels starting from 0, the Dchannel isnumbered 23 and the
logical voice port name isSerial-ContraHer ID:23, for example. Serial 1/1/1:23
The isdn overlap-receiving command allows the incoming numbers to be received digil-
by-digit and not en bloc.
lo define incoming calls asvoice-only, you configure isdn incoming-voice voice. This
will ^nd incomingcalls to the DSP resources.
1-120 Implemerting Cisco Voice Communications and QoS (CVOICE) v8 0
Fine-Tuning Analog and Digital Voice Ports
This topic describes how to fine-tune the analog and digital voice ports.
Fine-Tuning Voice Ports
Cross-Connect BetweenAnalogPort and Digital DSO
Cross-connect works for:
- Ports on the same module (NM-HD-2VE)
CASDSO groups consisting of one time slot
Maximum4 FXO/FXSports at the same time
The channel bank feature provides support fortheTDM cross-connect functionality between
analog voice ports and digital DSOs onthesame NM-HD-2VE using CAS. The cross-connect
works as a switch between the selected time slots on the Tl/El CAS trunk and an analog voice
interface.
These restrictions apply:
Theconfiguration for cross-connect must beon thesame network module.
Amaximum of four FXSor EXO portscan be cross-connected to a T1 interface.
ABRI-to-PRl or analog-to-BRI/PRI cross-connect cannot beconfigured; theonly
connection for cross-connect is analog-to-Tl/El CAS (dsO-group).
The DSO group must contain only onetime slot. Thesignaling type of theDSO group must
match that of the analog voice port.
if the channel bank feature is used for the Tl controller, the rest of the unused DSO group
cannot be used for fractional PRI signaling.
Fine-Tuning Voice Ports (Cont
Cross-Connect Configuration
controller el 0/0
dsO-group 0 timeslots 5 type
dsO-group 1 timeslots 8 type
voice-port 0/0:0
signal loop-start
voice-port 0/0:1
operation 2-wire
type 1
signal wink-start
connect connectl voice-port 0/1/0 el 0/0 0
connect connectl voice-port 0/1/1 el 0/0 1
Cross-Connect Configuration
PSTN
fxs-ground-start
e&m-fgd
lo establish a channel bank (cross-connect) connection between an analog voice port and a T
DSO. configure theconnect command inglobal configuration mode, fhe parameters of this
command include the analog voice port identifier, and the controller identifier with the DSO
number that should be cross-connected.
1-122 Implementing Cisco VoiceCommunications and QoS (CVOICE) v8 0
) 2010 Cisco Systems. Inc
Fine-Tuning Analog Voice Ports
This subtopic describes various timers that can be fine-tuned on analog voice ports.
Fine-Tuning Analog Voice Ports
Timing Parameters
tim eouts initial
- Howlong thedialtoneis presented beforethe first digil is expected
timeouls interdigit
- How long towaitforthe nextdigit beforethe number is considered
complete
timeoutsringing
How longa caler maylet the telephoneringwhenthere is noanswer
timing digit
DTMF digit signal duration
tming interdigit
DTMF interdigit duration
timing hookflash-inand hookflash-out
- maximum duration of a hookflash indication
hookflash is an indication bya callerto dosomethingspecial with the
call, such as transfer or place the cal on hold
Youcanset a numberof timers and timingparameters for fine-tuning the voice port. Ilere are
voice-port configuration parameters that you can set:
timeouts initial: Configures the initial digit timeout value inseconds. This value controls
how long the dial tone ispresented before the first digit isexpected. Ifthe first digit isnot
entered within this timeout, the reorder tone is sent. This timer value typically does not
need to be changed.
timeouts interdigit: Configures thenumber of seconds thatthesystem will wait between
caller-entered digits before sending the input tobe assessed. Ifthe digits are coming from
anautomated device and die dial plan isvariable-length, youcan shorten this timer sothat
the call proceeds without waiting the complete default of 10 seconds for the interdigit timer
toexpire. Ifthe timeout isexceeded before the complete DNIS iscollected, the call routing
will fail.
timeouts ringing: Configures the length of time that acaller may continue toletthe
telephone ring when there isno answer. You can configure this setting tobeless than the
defaultof 180secondsso that you do not tic up the voice port whenit is evident that the
call isnot going tobe answered. Ifthe timeout isexceeded, the caller receives the reorder
tone and the called line is released.
timingdigit: Configures the DTMF digit signal duration fora specified voice port. You
can use this setting tofine-tune a connection toa deviee that may have trouble recognizing
dialed digits. Ifa user ordevice dials too quickly, the digit may not berecognized. By
changing the timing onthe digit timer, you can provide for a shorter or longer DTMF
duration.
timinginterdigit: Configures theDTMF interdigit duration fora specified voice port. You
can change this setting toaccommodate faster or slower dialing characteristics.
timing hookflash-in and hookflash-out: Configures the maximum duration (in
milliseconds) ofahookflash indication. Hookflash is an indication by acaller that the caller
wishes to do something specific with the call, such as transfer the eall orplace the call on
hold. For the hookflash-in command, ifthe hookflash lasts longer than the specified limit,
the FXS interface processes the indication ason-hook. Ifyou set the value too low. the
hookflash mav be interpreted as ahang-up: if\ou set the value loo high, the handset has to
be left hung upfor a longer period toclear thecall. For the hookflash-out command, the
setting specifies the duration (in milliseconds) ofthe hookflash indication that the gateway
generates outbound. You can configure this tomalch therequirements of the connected
device.
Under normal use. these timers do not need tobe adjusted. There are twoinstances in
which these timers can beconfigured to allow more orless lime for aspecific function:
When ports are connected toa dev ice that does not properly respond to dialed digits
or hookflash
\\ hen the connected dev ice provides automated dialing
1-124 Implementing CiscoVoice Communications and QoS (CVOICE) v80 2010 CiscoSystems, Inc
Fine-Tuning Analog Voice Ports (Cont)
Timing Parameters Configuration
Present the dial tone for 15 seconds before the first digit is
collected.
Keepwaiting 15 seconds for the next digit.
Allow the phone to ring 4 minuteswhen there is no answer.
Hookflash indication may not exceed 0.5 seconds.
Fc:
voice-port 0/1/0
timeouts initial 15
timeouts interdigit 15
timeouts ringing 24C
timing hookflash -in 500
Timing Parameters Configuration
fhe example in the figure shows how timers can be tuned on avoice port toachieve these
requirements:
Present the dial lone for 15seconds before the first digit is collected.
Keep waiting 15secondsfor the next digit.
Allow the phone to ring 4 minutes when there is no answer.
Hookflash indication may not exceed 0.5 seconds.
Echo Cancellation
This topic describes how echo isgenerated ina telephone conversation and how the echo
cancellation feature works onCisco Iinified Communications gateways.
Echo Origin
Echois a signal that leaks from the Rx pathintothe Txpath.
* Typically, due to impedance mismatch at the two-wire to four-
wire hybrid connections.
Echo annoys callers ifabove amplitude and delay threshold.
Fcho is thesound of yourownvoice reverberating inthetelephone receiver while youare
talking. When timed propcrlv. echo is not a problem in a conversation; however, if the echo
interval exceedsapproximately 25 ms. it can be distracting lo the speaker. In Ihetraditional
telephonv network, echo is generally causedbv an impedance mismatch when the four-wire
network is convertedto the two-wire local loop.
The figure isan example of a two- tofour-wire hybrid circuit. Hybrid echo iscaused by an
impedance mismatch in the hybrid circuit. This mismatch causes the transmit (Tx) signal to
appear on the receive (Rx) signal.
fhe telephone eompanv (telco) usuallv applies its own port tuningtechniques to minimize
echo. Echo isconstant ina telco environment; however, low delay and low amplitude typically
make echo not an issue.
For echo to be a problem, all of the followingconditions must exist:
An analog leakage path between analog T.\ and R\ paths
Sufficient delav in echoreturn torechotobe perceived as annoying
Sufficient echo amplitude lobe perceived as annojing
Talker Echo
This subtopic explains what the talker echo is and how it occurs.
Talker Echo
*Signal leaks from Rx to Tx path at the remote end
TalkerEcho (MostCommon)
Talker echo occurs when the speech energy of atalker, transmitted down the primary signal
path is coupled into the receiving path from the far end (or tail circuit). Talkers then hear their
own voice dclaved bv the total echo path delay time. Ifthe echoed signal has sufficient
amplitude and delay, the result can be annoying to the customer and can interfere with the
normal speech process, Talker echo is usually adirect result of the two- to four-wire conversion
that takes place in die PSfN.
Listener Echo
1-128
Ihis subtopic discusses the listener echo.
-istener Echo
Signal leaks from Rx toTx path at both ends
Listener Echo (Less Common)
Listener echo occurs at the far end by circulating voice energy. Listener echo is generally
caused bv the two- and four-wire hybrid transformers (caused by the echo being echoed) The
voice ot the talker is echoed by the far-end hv brid. and when the echo comes back to the
listener, the hvbnd on the side ofthe listener echoes Ihe echo back toward the listener The
eflect is that the person listening hears both (he talker and an echo ofthe talker
2010 Cisco Systems
Echo Cancellation
This subtopic describes theecho canceller functionality.
Echo Cancellation
ITU-T G.168 Echo Canceller
Enabled by default.
Faces into the PSTN side of a voice gateway,
Captures and stores the outgoing voice signal.
Watches the Rx path for echo.
- Estimates the level
- Subtracts the original signal from the Rxsignal
When one side is silent, comfort noise is generated.
* Echo canceller coverage is the length of time that is stored in
memory and can be adjusted.
An echo cancelleris a tool that you can use lo controlecho. Anecho cancellerreducesthe level
of echo that leaks from the Rx path (from thegateway out into thetail circuit) intothe Txpath
(from ihetail circuit into the gateway). From theperspective of theecho canceller ina voice
gateway, the Rx signal isa voice that comes across the network from another location. The Tx
signal isa mixture ofthe voice call inthe other location and the echo ofthe original voice.
which comes fromthe tail circuit on the initiating end and is sent to the receiving end.
Echo cancellers face into the PSTN tail circuit and eliminate echo in the tail circuit. The echo
canceller intheoriginating gateway looks out intothetail circuit. It is responsible for
eliminating theechosignal from the initiating Tx signal andallowing a voice call togothrough
unimpeded. By design, echo cancellers are limited by thetotal amount of time that they wait for
the reflected speech to bereceived, which isknown as anecho tail. Theecho tail is normally
32 ms.
Echo cancellation is implemented in the DSP firmware on Cisco voice gateways andis
independent of other functions implemented inthe DSP, suchas the DSP protocol and
compression algorithm. Invoice packet-based networks, echo cancellers arebuilt into thelow-
bit-rate coder-decoders (codecs) and operate on each DSP.
Echo Canceller Parameters
1hissubtopic describes theechocanceller parameters.
xho Canceller Pan
ERL (echo return loss) Represents the reduction of returning echo
(larger is better)
ERLE (ERL enhancement) Additional echo loss from canceller
ACOM= ERL + ERLE (larger is better)
Use output attenuation and input gain to tune ERLto at least 6 dB
Echo Cancel Coverage
(Tail Length)
ERLE
Echo Canceller Operation
Anecho cancellerremoves the echo portion of thesignal comingout of the tail circuit and
headed into the WAN. To do this, the echo canceller learns the electrical characteristics of the
tail circuit and forms its own model of the tail circuit in its memory, and createsan estimated
echo signal based on the current and past R\ signal. The echo canceller subtracts the estimated
echo from the actual T\ signal coming out of the tail circuit. The quality of the estimation is
continuous!} improved because the echo canceller monitors the estimation error.
Echo Canceller Components
A typical echo canceller includes two components: a convolution processor and a nonlinear
processor (Nl.P),
The convolution processorfirst stage captures and stores the outgoingsignal toward the far-end
hvbrid, Ihe convolution processor then switches to monitoring mode and. when theechosignal
returns, estimates the level of the incoming echosignal, and subtracts the attenuated original
voice signal from the echo signal.
The time that it takes to adjust the level of attenuationto the original signal is called the
convergence time. Because the convergence process requires that the voicesignal be stored in
memory, theecho cancellerhas limited coverage of tail circuit delay, normally 64 ms. 96 ins.
and up to 128ms. After convergence, the convolution processor provides about 18dB of echo
return loss enhancement (ERIE). Becausea typical analog phone circuit provides at least 12
dB of echo return loss (ERL) (that is. the echo path loss between the echo canceller and the far-
end hybrid), the expected permanent l-.RL of the converged echo canceller is about 30 dB or
preater.
1-130 Implemenling Cisco Voice Communications and QoS (CVOICE] v8.0
"Ihe second component is the NLP. In single-talk mode, that is, when one person is talking and
the other issilent, the NLP replaces the residual echo atthe output ofthe echo canceller with
comfort noise based onthe actual background noise of the voice path. The background noise
normally changes over the course ofaphone conversation, so the NLP must adapt over time.
"Ihe NLP provides an additional loss ofat least 25 dB when activated. In double-talk mode, the
NLP must bedeactivated, because it would create a one-way voice effect byadding 25 to30
dB of loss in only one direction.
Configuring Echo Cancellation
fhis subtopic explains how to fine-tune the echo canceller.
"iqurn
router(config-voiceport)#
echo-cancel coverage {8 16 24 32 48 64}
* Adjusts the time coverage of the echo canceller
Available options and default value differ per platform and
software version
router(config-voiceport)#
(no] echo-cancel enable
Disables or re-enables echo canceller
Default, enabled
Echo cancellercoverage (alsoknown as tail coverage or tail length)is the length of time that
the echo canceller stores its approximation of an echo in memory. An echo canceller can
eliminate the maximum echo delay.
Theecho cancellerfaces into a static tail circuit withan inputand an output. If a wordenters a
tail circuit, the echo is a seriesof delayed and attenuated versions of that word, depending on
the numberof echosources anddelays associated withthem. After a certainperiod, nosignal
comesout, fhis time periodis known as the ringingtimeof the tail circuitthe time required
for all of the ripples to disperse. To fully eliminate all echoes, the coverage of theecho
canceller must be as long as the ringing time of the tail circuit. Use the command echo-cancel
coverage timeto set the tail coverage. The available lime options and the default value differ
per platform and software version.
To change the threshold at which the gateway will be able to delect echo, use the command
echo-cancel erl worst-case {6 j 3 | 0]. Eorexample, if you have a worst-case ERL of 6 (echo-
cancel erl worst-case 6). when you speak into the phone you can expect at least 6 dB of
attenuation on the signal bv the time it gels backlo the original source(echo). Ingeneral,you
do not need to change this value fromthe default of 6. Selling the worst-case ERE does not
di-ectly modify the inbound or outbound signals, fhis is purely a configuration parameter for
the echo canceller to help it distinguish between echo and a newsignal.
You can disable and re-enable the echo canceller using the echo-cancel enable and no echo-
cancel enable commands in the voice port configuration mode. The canceller is enabled bv
default.
1-132 Implementing Cisco Voice Communications and QoS (CVOICE) v8 0 '2010 Cisco Systems, Inc
Verifying Analog and Digital Voice Ports
This topic describes how to verify the operations ofanalog and digital voice ports.
Verifying Voice Ports
Command Overview
Command
showvoice portfsJbeportl
summary]
show controllers bri slot/port
showcontrollers t1 slot/port
showcontrollers e1 slot/pod
showvoice call summary
show call active voice
showcall history votes
Description
Displaysconfiguration informationabout a Specific
voice port or a summary of all voice pom
Displays information about the specified voice port
Verfies the call status foral voice ports
Displays the contents ofthe activecall table
Displays tie contentsofthe caShistory table
The table inthe figure provides an overview ofthe most useful commands when verifying the
operations of analogand digital voice ports.
The show voice port summarycommand identifies the port numbers of voice interfaces
installed on the voice gateway.
The show voice port command verifies voice-port parameter settings.
The show controllers tl/cl/bri commandwith the slot/port options verifies the controller
settings.
The show voicecall summary command verifies thecall statusfor all voice ports.
The showcall active voice and showcall history voice commands displaythe current and past
call activity.
show voice port summary Command
This subtopic provides anexample output of the show voice port summarycommand.
show voice port sui
Route r# a how voire port summar>
in OUT
PORT CB SIG-TYPE ADMIN CFER STATUS STATUS EC
0 :17 IS fxo-ls do* Q down idle on - hoo k
y
0 :16 19 fio-ls up do cm idle on-hook
Y
0:15 20 fio-ls up dorm idle on-hook
y
0 : 2C 21 fxo-l! up dorm idle on-hook
y
0:21 2J xo-ia up dona idle on -hoo k y
0 :22 23 fxo-ls up derm idle on -hoo k
Y
0:23 24 em-imd up dorm idle idle Y
1/1 X3-ls up dorm on-hook idle y
1/2 xs-1b up dorm on - hoo k idle
y
1/3 etm-imd up dorm idle idle
Y
1/4 e im- imd up dorm idle idle Y
1/5 fxo-la up dorm idle on -hoo k
y
1/6 IXO-1S up dons idle on -hoo k
y
The figure displavs a sample output of theshow voice port summary command, ll displays the
available voiceports, including the slot numbers, if they are carried over a digital voiceport.
Thesummary includes thesignaling type configured on theport, administrative andoperation
state, in and out status, and the echo cancellation status.
Verifying Analog Voice Ports
This subtopic provides an example output ofthe show voice port command.
Verifying Analog Voice Ports
show voice port
Router* show voice port
DSO Group 1:0 - 1:0
Type of voicaPort is CAS
Operation State in DORMANT
Administrative State is UP
No Interface Bonn Failure
Description ii not Hat
Hoiae Regeneration i enabled
Hon Linear Processing ia enabled
Music On Bold Threshold is Set to -3B dBm
In Gain is Set to 0 da
Out Attenuation is Set to 0 dB
Echo Cancellation is enabled
Echo Cancel Coverage is set to B ms
Flayout-delay Mode is set to default
Playout-delay Hominel is set to SO me
Playout-delay Maximum in set to 200 me
Connection Mode is normal
Connection Number 1b not set
The figure displays asample output ofdie show voice port command. Itprovides information
about the voice port type. DSO group, and operation and administrative status, and offers
infonnation about the echo canceller and other parameters.
This subtopic provides an example output of Iheshow controller command.
show controller
Router* show controller Tl 1/0/0
Tl 1,'0,'C la up.
Applique type is Channelized Tl
Cablelength la long gain36 Odb
No alarms detected.
alarm-trigger ia not set
Framing ib ESP, Line Code is B8ZS, cloc* Source is Line.
Data in current interval (180 seconds elapsed):
0 Line code violations, 0 Path Code Violations
0 Slip Sees, 0 Pr Loss Sees. 0 Line Err Sees, 0 Degraded Ming
0 Errored Sees, 0 Bursty Err Sees, D Severely Err Sees, 0 Unavail
T'te figure displavs a sample output of theshowcontrollercommand. Thecommand provides
infonnation abotit the controller tvpe and status, the framing, line coding, clocksource, and
reports of transmission errors.
1-136 Implementing Cisco Voice Communications and QoS (CVOICE) v8.i
Verifying Voice Ports (Cont)
show voice call summary
Router* sbo" voice call sunnuary
PORT CODBC VAD VTSP STATE VPK STATE
1/015,1 g72r8 y 3 CONNECT S ISP COMNECT
1/015.2 gT29r8 y S CONNECT S TSP CONNECT
1/015.3 g729r8 y S COMNECT S TSP CONNECT
1/015.4 g729r8 y S CONNECT S TSP CONNECT
1/015. S g729rfl y S CONNECT S TSP CONNECT
1/015.6 g729r8 y S CONNECT 3 T3P CONNECT
1/015.7 g729rB y S CONHBCT S TSP CONNECT
1/015.B gllStB y S CONNECT S TSP CONNECT
1/01S.9 g729rB Y S CONNECT S TSP CONNECT
1/015.10 g729rB y S CONNECT S TSP CONNECT
1/015.11 g729r8 y S CONNECT S TSP CONNECT
1/015.12 g729t8 y S CONNECT STSFCONNECT
show voice call summary Command
The figure displays a sample output of the showvoicecall summary command. It provides
information about the current usageof the available voice ports, including the time slots of
logical voiceports, and VoIPparameters (codecand voice activity detection [VAD]) that are
applied tothecallswhen thecircuit-based call is converted to VoIP. The command shows the
state of the Voice Telephony Service Provider (VTSP).
) 2010 Cisco Systems, Inc. Introduction to Voice Gateways 1-137
Verifying Voice Port!
Router* show call active \ oice
GENERIC:
SetupTlme = S4523746 ms
mdex*.148
PeerAddresa=*#73072
PeerSubAddress=
Peerld=70000
Peerlflndex.3-'
LogicalIfIndex=0
Connect Time^9452404 3
DisconectTine=94 546241
Ca HQrigin= 1
ChargedUnitsÔ
InfoType=2
Transw :Packets = 625 1
TransnitBytes=12 5023
ReceivePac)tets=3300
ReceiveBytea=66G00
show call active voice Command
The figure displav s a sample output of the show call active voice command, fhe command
ofiers detailed informationabout the current calls, includingthe start time and duration. If one
of the call legs is VoIP, it shows the peer II' address and traflic statistics.
1-138 Implementing Cisco Voice Communications and QoS (CVOICE] v8.0 2010 Cisco Systems, Inc.
show call history voice
Routert show call hietory voice
GENERIC:
SetupTime.54B93250 ma
index* 4 50
PeerAddreBB.##5225 8
PeecSubAddreeB-
peetId-50000
peerltlndex-35
LogicallfIndex.0
DiscoonectCeuBe-10
DisconnectText-normal call clearing.
ConnectTlme-94 8937 80
DisconectTiae.95015500
CallOrigin-1
Chargedunits-O
InfoType-2
TranSmitPc>ietB-32 2S8
TransmitBytes-645160
Receivers, chets-20061
BeceiveBytee-4 0122 0
show call history voice Command
The figure displavs a sample output of the show call history voice command. This command
provides informationequivalent to the show call active voice command, except that it
describes calls that occurred in the past.
i 2010 Cisco Systems, Inc. Introduction to Voice Gateways 1-139
Verifying Voice Ports (Cont.)
test Command OvervievV
Tests detector-relaled functions on a voice
port
.HHCommand
test voice port slot/subunit/put
detector
test voice port sloVsubunit/port
inject-tone {local | network}
1njects a test tone into a voice port
test voice port slot/subunit/port
loopoack{local | network [disable)
Performs loopback testing on a voice port
test voice port slot/subunit/port relay Tests relay-related functions on a voice port
test voice port slot/subunit/port
switch (fax | disable)
Forces a voice port into fax mode
csim start xMX Initiates simulated calls
A call must be established prio to all testing commands, except csim.
test voice port and csim Commands
The table presents an overview of the test voice port commands, and the csim command that
can be used to simulate a call when troubleshooting gateway operations.
Implementing Cisco Voice Communications and QoS (CVOICE) v8.0 2010Cisco Systems. Inc
Voice Port Testing Example
routert
test voice port inject-tone local 500hz
Inject a 500-Hztest tone intothe voice port
1/0/1
Voice Port Testing Example
Voice port test commands are helpful when troubleshooting a problem. The figure shows how
to test a voice port by injecting a 500-Hz tone into it.
) 2010 Cisco Systems, Inc Introduction lo Voice Gateways 1-141
Summary
Ihis topic summarizes the kev points that were discussed in this lesson.
Summary
Digitalvoice ports use TDM to carry multiple voice channels over a
single circuit
FXS ports connect aateways to end-user equipment, FXOports to
CO switches or PBXs, and E&Mports to PBXs.
Analog port settings define country-specific voice parameters,
signaling type, and in case of E&M, circuitry.
Digitalvoice ports emulate analog signaling (CAS) or use CCS
ISDN uses CCS on BRI and PRI interfaces.
Digital voice ports are defined by creating DSO groups on the
T1/E1 controller and configuring their signaling type
ISDN PRI interfaces are created by defining a pri-group on the
T1/E1 controller
Tunable voice port parameters include cross-connects, timing, and
comfort noise
Echo cancellation is enabled by default but its time coverage can
be tuned
Voice port and call parameters can be viewed
1-142 Implementing Cisco Voice Communications and QoS (CVOICE) v8.0 2010 Cisco Systems, Inc
Lesson 4
Understanding DSP
Functionality, Codecs, and
Codec Complexity
Overview
This lessonexplains the compression schemes that you can use to transport voiceusingvarious
coder-decoders (codecs), and the implications of these compression schemeson bandwidth
usage. To understand the bandwidth issuesthat youwill encounter, you must be able to
calculate the amount of bandwidth that a VoIP call will consume.
Furthermore, this lesson discusses the digital signal processors (DSPs) that convert analog and
digital voice signals into VoIP traffic.
Objectives
Uponcompleting this lesson,you will be able to define DSPsand codecs, explaindifferent
codec complexities, and choosethe appropriate codec for networkuse. This abilityincludes
being able to meet these objectives:
Explain voice codecs and their major features
List voice quality evaluation methods and explain how they are applied to voice codecs
Describe howthe packet rateand protocol overheadimpacts the total per-call bandwidth
Explainvarious types of DSPs. DSP functions, and how DSPs are used in voice gateways
Describe codec complexities
Configure a DSP for voice termination at a voice gateway
Monitor and verify operation of DSPs
Voice Codecs
This topic describes thevoice codecs supported byCisco Unified Communications gateway;
Voice Codecs
Codec Bandwidth
j0HHIflH2!3IHIMIItiii
G.711/G.722 64
G 726r32 32
G.726r24 24
G726M6 16
G.728 16
iLBC' 15.2, 13.3
GSM Full Rate (GSM-FR) 13
G729 8
G.723r63 6.3
G 723r53 5.3
iLBC - Inte-net Lew Brtrate Codec
A codec is a DSP software algorithm that compresses and decompresses speech or audio
signals. There are mam standardized codecs used in VoIP networks, including pulse code
modulation (PCM), adaptive differential PCM (ADPCM}, low-delay eode-exeitcd linear
prediction (LD-CELP). and conjugate-structure algebraic-code-excited linear prediction (CS-
ACELP).
Following is a list of major codecs supported by the Cisco IOS gateways:
G.711: International standard for encoding telephone audio on a 64-kb/s channel. It is a
PCM scheme operating at an 8-kH/ sample rate, with 8 bits per sample. With G.711, the
encoded voice is already in the correct format for digital voice delivery in the public
switched telephone network (PSTN). It is widelv used in the telecommunications field
because it improves the signal-to-noise ratio without increasing the amount of data.
There are two subsets of the G.711 codec:
mu-law: Used in North American and Japanese phone networks
a-law: I ised in Europe and elsewhere around the world
Both mu-law and a-lavv subsets use compressed speech carried in 8-bit samples. They use
an 8-kllz sample rate with 64 kb/s ol'storage.
G.722: ITU ADPCM standard speech codec that provides 7 kHz of wideband audio at data
rates from 48 to 64 kb/s. Of two variants. G. 722.1 and G.722.2, G.722.2 ofiers better
compression as well as the ability to quickly adapt changing network conditions.
Bandw idth is automatically conserved when network congestion is high. When congestion
returns to a normal level, a lower-compression, higher-quality bit rale is restored.
1-144 Implementing Cisco Voice Communications and QoS (CVOICE) v8.0 2010 Cisco Systems Inc.
G.726: ITUADPCM schemethat encodesanalogvoicesignals intodigital signals by
predicting future encodings based on the immediate past. The adaptive feature ofADPCM
'*mr reduces the number of bits per second that the PCM method requires to encode voice
signals.
G.728: ITU 16-kb/s LD-CELP variation of CELP voice compression.
W G.729: CS-ACELP compression to code voice into 8-kb/s streams. There are two
variations of thisstandard (G.729 andG.729 Annex A[G.729A]) thatdiffermainly in
computational complexity. There areextensions, which also provide 6.4-kb/s (Annex D)
*^ and 11.8-kb/s (Annex E) rates for marginally worse and better speech quality, respectively.
G.723: Dual-rate speech coder formultimedia communications. Thiscompression
technique can beused for compressing speech or audio signal components at a very low bit
rate as part of theH.324 family of standards, fhis codec has twobit rates associated with it:
r63: 6.3 kb/s, using24-bytc frames andthe Multipulse EPCwith Maximum
Likelihood Quantization (MPC-MLQ) algorithm
^ r53: 5.3 kb/s, using 20-byte frames and the ACELP algorithm
fhe higher bitrate is based on MPC-MLQ technology and provides a somewhat higher
quality of sound. The lower bit rate is based onCELP and provides system designers with
"^ additional flexibility.
Global System for MobileCommunications Full Rate (GSM FR) codec: Usesa frame
siz.e of 20 ms and a bit rate of 13 kb/s. It is a Regular Pulse Excitationwith Long-Term
'. Prediction (RPE-LTP) coder. It is required for Voice Extensible Markup Language
(VoiceXML) scripts that canfunction as theuserinterface for a simple voice-mail system.
Thiscodec supports theCisco infrastructure andapplication components required for
^v service providers todeploy unified messaging applications.
r Internet Low Bitrate Codec (iLBC): Designed for narrowband speech, this codec results
ina payload bit rate of 13.33 kb/sfor 30-ms frames and 15.20 kb/sfor20-ms frames, fhe
... . algorithm is a version of block-independent linear predictive coding, with thechoice of
^ data frame lengths of20and 30ms. This codec enables graceful speech quality degradation
in the case of lost frames, which occurs in connection with lost or delayed IP packets.
*, The figure provides a list of the codecssupported on CiscoUnified Communications gateways
teg andtheVoIP bandwidth thatthey use. Thebandwidth doesnot include anypacketization
overhead and is often referred to as raw bandwidth.
2010 Cisco Systems, Inc. Introduction JoVoice Gateways 1-145
Voice Codecs (Cont.)
G.729/G.729A/iLBC Comparison
* G.729/G.729A
Both ITUstandards, 8-kb/s. same compression delay (10
to 20 ms)
G.729Aless complex and processor-intensive, slightly
worse quality than G.729
The Annex B variant can be applied to either codec
* Adds VAD and CNG
iLBC
Similar complexity but better quality than G.729
Supports two fixed-bit-rate frame lengths:
* 13.3 kb/s with an encoding frame length of 30 ms
* 15.2 kb/s with an encoding frame length of 20 ms
Supported on Cisco Unified IP phones and Cisco gateway
dial peers
G.729 Variants
iLBC
(i.729 is commonly used for high-quality 8-kb/s voice. G.729 is a high-complexity, processor-
intensive compression algorithm.
G.729. G.729 Annex A (G.729A). G.729 Annex B (G.729B). and G.729 Annex A with Annex
B (G.729AB) are variations of CS-ACELP. There is little difference between the ITU
recommendations for (i.729 and G.729A. All of the platforms that support G.729 also support
G.729A.
Although G.729A is also an 8-kb/s compression algorithm, it is not as processor-intensive as
G.729. G.729A is a medium-complexity variant of (i.729 with slightly lower voice quality and
is more susceptible to network irregularities such as delay, variation, and "landeming."
Tandeming causes distortion that occurs when speech is coded, decoded, then coded and
decoded again, much like the distortion that occurs when a tape is copied several times.
The Annex B variant of G.729 adds support for voice activity detection (VAD) and comfort
noise generation (CNG).
fhe following G.729 eodec combinations are interoperable:
G.729 and G.729A
G.729BandG.729AB
iLBC was developed in 2000 to meet the needs of the VoIP industry. The objective was to
develop a codec that is royalty free, designed specifically for packet communication, and that
offers high voice quality for both clean and packet-loss conditions, and to bring the codec
forward to ditTerent standardization bodies for interoperability compliance.
iLBC includes the following features:
It is royalty free.
It has belter quality than G.729.
It has similar computational complexity as G.729.
The iLBC codec supports twobasicframe lengths:
A bit rate of 13.3kb/s with an encoding frame length of 30 ms.
A bit rate of 15.2kb/s with an encoding frame length of 20 ms.
The iLBC codec enables graceful speech quality degradation inthecase of lost frames,
which occurs in connections with lost or delayed IP packets.
It has betterquality than G.729 withbetterrobustness topacketloss.
It is supported on CiscoUnified IPphones.
It issupported onCisco Unified Communication gateways, and inVoIP dial peers using
H.323 and Session Initiation Protocol (SIP).
i 2010 Cisco Systems. Inc Introduction to Voice Gateways 1-147
Voice Codec Packet Rates and Payload Sizes
This subtopic explains thepacket raleandpayload sizeoptions available for voice codecs.
Voice Codec Packet Rates an
d Payload
Sizes
Codec Packets per Second (PIS)' !Paylcs(i;Size;(fpg
G 711 33 240
G 711 50 160
G 726f32 33 120
G 726r32 50 SO
G 726M6 25 80
G 726M6 50 40
G729(aj3,ab) 33 30
G729(a.6,at 50 20
iLBC 33 50
iiec 50 38
G 723rB3 17 46
G 723r63 33 24
G.723r53 !7 40
G 723r53 33 20
ow-band*pdth codec s produce samples al lower rate
Voice sample si/e is a variable thatcanaffect thetotal bandwidth used. Avoice sample is
defined as thedigital output from a codec DSP that is encapsulated ina VoIP packet. Cisco
uses DSPs that sample output based ondigitization of 10msof audio. Cisco voice equipment
encapsulates a minimum of 20 ms of audioin eachpacket. Youcan applyan optional
configuration command to the dial peer lo varythe numberof samplesencapsulated.
Thetable inthe figure illustrates various codecs andtheirpacket-per-second rates andpayload
sizes. The largerthe pav load, the largerthe packet, but the fewerpackets have to be sent, which
reduces bandwidth overhead.
fhe formula for computing the codec bandwidth is as follows:
packet_ratio * paylcad_size * 8 (bits)
Most packet-per-second rates are rounded for simpler presentation. 33 p/s should more
precisely be written as 33.3 p/s and 17p/s should correctly read 16.6p/s.
Examples are:
Bandwidth of G.7 II with 33.3 p/s is; 33.3 packets * 240 B* 8 (b/B) = 64,000 h/s
Bandwidth of G.71 1with 50 p/s is; 50 packets * 160 B * 8 (b/B) = 64.000 b/s
Bandwidth of G.729 with 33.3 p/s is: 33.3 packets * 30 B * 8 (b/B) = 8000 b/s
Bandwidth of (i.729 with 50 p/s is: 50 packets * 20 B * 8 (b/B) = 8000 b/s
Evaluating Quality of Codecs
This topic describes the methodology toevaluate the quality ofvoice codecs.
Voice Quality Evaluation
Test Methods
Mean opinion score (MOS):
Defined in ITU-T Recommendation P.800
Results in subjective measures
Scores from 1 (worst) to 5 (best); 4.0 is business quality
Perceptual Evaluation of Speech Quality (PESQ):
Automated assessment of the speech quality as experienced by users
Successor of Perceptual Speech QualityMeasurement (PSQM)
ITU-T recommendation P.862 (Feb 2001)
Woridwide applied industry standard forobjective voicequaity testing
PESQ results princpaly modelmean opinion scores (MOSs)
Perceptual Evaluation ofAudio Quality (PEAQ):
Automated assessment of speech and other audio types
Patented and available under license
PEAQ results principaly model MOSs
Mean Opinion Score
Mean opinion score (MOS) isa scoring system for voice quality. An MOS score is generated
when listeners evaluate prerecorded sentences thataresubject to varying conditions, such as
compression algorithms. Listeners then assign values tothesentences based ona scale from l
to 5. where I is the worst and 5 is the best.
The test scoresarc then averaged to a composite score. The test results are subjective, because
they are based onthe opinions ofthe listeners. The tests arc also relative, because ascore of3.8
from onetest cannot bedirectly compared to a scoreof 3.8from another test. Therefore, a
baseline for all tests must be established so that the scores can be normalized and compared.
Perceptual Evaluation of Speech Quality
Perceptual Evaluation of Speech Quality (PESQ) isa family of standards comprising a test
methodology for automated assessment of thespeech quality as experienced bya userof a
telephony system. Dctlned as ITU-T recommendation P.862 (Feb 2001). it isa worldwide
applied industry standard for objective voice quality testing, used byphone manufacturers,
network equipment vendors, andtelcooperators. PESQ cantakeintoaccount codec errors,
filtering errors, jitterproblems, and delay problems thataretypical ina VoIP network. PESQ
scores range from 1(worst) to4.5 (best), with 3.8considered toll quality thatcan bemapped to
MOS scores. PESQ replaces its predecessor, Perceptual Speech Quality Measurement (PSQM).
Perceptual Evaluation of Audio Quality
Perceptual Evaluation ofAudio Qualitv (PEAQ) is astandardized algorithm for objectively
measuring perceived audio quality, not onlv speech. Defined as ITU-R Recommendation
US. 1387. it utilizes software tosimulate perceptual properties of thehuman ear andthen
integrate multiple model output variables into a single metric. PE.AQ characterizes the
perceived audio qualitv assubjects would doina listening test, PEAQ results principally model
mean opinion scores (MOSs) that cover a scalefrom 1(bad) to5 (excellent).
fhe PEAQ technologv isprotected by several patents and isavailable under license, together
with the original code for commercial applications. However, free, unvalidated PEAQ model
implementations exisl.
Voice Quality Evaluation
Test Methods Comparison
Test method
End-to-end
packet bss test
Subjective
Inconsistent
End-to-end jitter Inconsistent
test
Measurement
subject
Test Methods Comparison
Voice and
other audio
Objective Objective Objective
No Yes Yes
No Yes
Voice
Yes
Voice and
other audio
The table inthe figure summarizes the key features of the described methods: meanopinion
score. Perceptual Evaluation of Speech Qualitv. Perceptual Evaluation of AudioQuality, and
thepredecessor of PESQ. Perceptual Speech Quality Measurement. Inessence. PSQM. PESQ,
and PEAQ provide an objective methodology that can be mapped to the subjective MOS
model. Thecurrent standards. PESQ and PEAQ. include a complete range of factors that would
be alsoconsidered bv a subjective test, PEAQ differs from PESQ mainly in that it is also used
to evaluate other audio types.
Implementing Cisco Voice Communications and QoS (CVOICE) v8 0 )20I0 Cisco Systems. Inc
Codec Quality
This subtopic provides the MOS scores for the common codecs.
Codec Quality
Codec
I Bandwidth [kb/s] | MOS*
G.711 64 ;4.3
G.726r32 32
3.8
G.726r24 24 3.75
G.726r16 16
3.7
G.728
16 3.75
iLBC
15.2 4.14
GSM Full Rate
13 3.5
G.729 8 3.92
G.729a
8
37
G.723r63 6.3 3.7
G.723r53 5.3 3.65
MOS values under ideal network conditions no packet loss, lowdelay, and no jitter
The table in the figure provides the average MOS scores for most typical codecs. These values
represent MOS scores under ideal network conditionsno packet loss, low delay, and no jitter.
TheMOS values measured under heavy network load will differ from thevalues shown inthis
table.
Evaluating Overhead
fhis topic describes how the packetization ratio impacts eall bandwidth.
Evaluating Overhead
Bandwidth Calculation
The table lists Layer 3+bandwidth percall, excluding overhead.
More accurate bandwidth percall calculation would include Layer
2 overhead.
Packetization
Period
Voice
Payload
160 Byte j 50 80 kb/s
74 kb/s
j 50 24 kb/s
| 33 18 kb/s
Bandvudth per call =
(Voice payload +Layer 3*overhead +tayer 2overhead] *packets per second *8 bits/byte
The packetization periodand the related voicepayload size affect the rawvoice bandwidth. The
table inthe figure illustrates the most common codecs with selected packetization periods,
payload sizes, packet ratios, and the resulting voice bandwidth, including theoverhead
introduced by Laver3 and above. The longer the packetization period is, the larger the sample
size is. and the lower the I.aver 3^ voice bandwidth is.
To compute the total call bandwidth, the additional Layer 2header musl he considered, using
the following formula:
EW_per_call = (\'oice_payload + L3+_overhead + L2_overhead) *
Packet ratio) * 8 bits/byte
1-152 Implementing Cisco Voice Communications and QoS (CVOICE] v8 0
2010Cisco Systems, Inc.
This subtopic describes the overhead introduced by various protocols and encapsulations.
Evaluating Overhead (Cont)
Layer 2 and Layer 3+Overhead
Layer 2 Headers [Bytes]
802.3 Ethernet
802 1Q Ethernet
PPP
MuitilinkPPPwith Interleaving
Frame Relay
Frame Relay with FRF.12
18
18+4
6-9
13
6
Layer 3 + Headers [Bytes]
VPN Headers Bytes]
IP
UDP
RTP
20
12
ESP
GRE/L2TP
MPLS label
50-57
24
4
Several factors must be included incalculating theoverhead of a VoIP call. Layer 2 and
security protocols significantly add tothe packet size.
Data-Link Overhead
Asignificant contributing factor to bandwidth isthe Layer 2protocol that isused totransport
VoIP. VoIP alone carries a40-bytc IP, User Datagram Protocol (UDP) and Real-Time
Transport Protocol (RTP) header. The larger the Layer 2overhead, the more bandwidth that is
required to transport VoIP:
IEEL 802.3Ethernet: Carries 18bytesof overhead; 6 bytesfor souree MAC. 6 bytes for
destination MAC. 2bytes for type, and 4bytes for cyclic redundancy check (CRC).
IEEE 802.1Q Ethernet: In addition tothe 802.3 overhead, there isa 32-bit 802. IQ header
that carries, amongothers, a 12-bitVLAN ID.
PPP: Carries 4to8 bytes of overhead. ThePPP header includes a 1- to2-byte flag to
indicate the beginning orend ofaframe (in successive frames only one character is used).
Qto 1address byte. 0 to 1control byte, 1- to2-byte protocol field, and 2bytes for CRC. If
both PPP peers agree to perform address and control field compression during Link Control
Protocol (LCP) negotiation, the control and address fields are not included. Ifboth PPP
peers agree to perform protocol field compression during LCP negotiation, the protocol
field is one byte.
Frame Relay: Carries 6bytes ofoverhead: 2bytes ofheader, 2bytes oftrailer (CRC). and
2 bytes of flags.
Frame Relay Fragmentation Implementation Agreement (FRF.I2): Inaddition lothe
Frame Relay overhead, there isa2-byte FRF.12 subheader that includes 4bits of flags and
a 12-bit sequence number tofacilitate reassembly at theremote end.
IP and Upper Layers Overhead
The IP and transport layers also have overhead to contribute to the size ofthe packets:
IP: Adds a 20-bvte header
LDP: Adds an 8-byte header
RTP: Adds a 12-b>te header
VPN Overhead
VPN encapsulation adds an additional overhead tothe VoIP packets:
Encapsulating Security Payload (ESP): Adds typically a50- to 57-byte overhead. Two
variables affect the ESP overhead: cipher block size and the authentication algorithm. The
tvpical block size iseight octets, but Advanced Encryption Standard (AES) works with 16-
octet block sizes. The block size influences the size of the initialization vector field, which
isthe same asthe block size plus the padding overhead, which can be up loblock size
minus oneoctet. The authentication algorithm yields different fingerprint sizes:
Message Digest 5 (MD5): 16 octets
Secure Hash Algorithm I (SHA-1): 20 octets
Secure Hash Algorithm 192 (SUA-192): 24 octets
Secure Hash Algorithm 256 (SHA-256): 32 octets
Generic Routing Encapsulation (GRE), layer 2 Tunneling Protocol (L.2TP): Adds a
24-bvte header.
Multiprotocol Label Switching (MPLS): Adds a 4-byte header for every label carried in
the packet. Alabel stackmay include multiple labelsin an MPLS VPN or traffic
engineering env ironment.
1-154 Implementing CiscoVoice Communications and QoS (CVOICE) v80 2010CiscoSystems, Inc.
Evaluating Overhead (Cont.)
Bandwidth Calculation Example
Example: Layer 3+, G.711 over Frame Relay, 50 Packets per
Second
Sandwidth per call _, ....
=(Voice paylcad +Layer 3OH +Layer 2OH) *packets per second x8bits/byte
=(160 +40+6) bytes *50 pps *8 bit/byte
-82.400 b/s-82 4 kb/s
Bandwidth Calculation Example
The example calculates the total bandwidth for aG.711 voice call with 50 p/s carried over a
Frame Relay network.
Tocompute thetotal call bandwidth, this formula is used:
Bandwidth_per_call - <Voice_payload + Layer 3_overhead + Layer
2_overhead) * PACKET_ratio) * 8 bits/byte
For the specified call, the bandwidth computes tothe following:
Bandwidth_per_call - {160 +40+6) - 50) * 8 bits/byte = 82,400 b/s
82.4 kb/s
Per-Call Bandwidth Using Common Codecs
This subtopic summarizes the total bandwidth for common codecs in various network typt
ir-Cali Ban
Codec Voice PPS i 0n,y Layer j Call over
Payload 3+ Frame Relay
G711 : 160 bytes
G.711 ^.240 bytes
G.729 , 20 bytBS
G.729 [ 30 bytes
_50_
33
50
33
_80kb/s
74.66 kb/s
24 kb/s
82.4 kb/s
76.27 kb/s
26.4 kb/s
' 18.66 kb/s ! 20.27 kb/s
87.2 kb/s
79,47 kb/s
31.2 kb/s
23,47 kb/s
'Ihe table inthe ligure includes the total eall bandwidth used by themostcommon codecs. It
lists the Lav er3t bandwidth and the total call bandwidth over 802.3 Hthemet and Frame Relay
networks. The La\er3- bandwidth takes into account the voice payload and II', UDP. and RTP
overhead. Ihe802.3 Ethernet and Frame Relay bandwidths consider the additional Layer 2
overhead. Thetable is produced usingthe formula introduced earlier.
1-156 Implementing CiscoVoice Communications and QoS (CVOICE) v8 0
Digital Signal Processors
This topic describes DSPs, their functions, and usage.
Example
Digital Signal Processors
Overview
PSTN
PSTN
Analog or
Digital
Analog or
Digital
Speech
DSPs
IP Packets
IP Packets
IP Packets
DSPs are silicon chips that perform specialized voice functions in real time. The DSPs inCisco
Unified Communications systems convert from onevoice format toanother. Cisco voice
gateways use the DSPs to terminate analog or digital circuits and convert the voice information
to VoIP packets, and vice versa.
Cisco Unified IP phones also contain DSPs, which convert voices originating on receivers of IP
phones lo IP packets. Conversely, voice packets that are received by IP phones as data are
converted back toananalog format and sent tothetransmitter (earpiece), which plays the
voices.
The terms -DSP" and "media resource" can be used interchangeably.
When an inbound call to a channel associated signaling (CAS) Tl that terminates on a Cisco
Unified Communications gateway isanswered, the DSP chip converts the inbound voice data
and sends it asVoIP packets toward the destination. The DSP converts the outbound voice that
is contained inthe VoIPpackets to the CASTl digital format,
DSP Functions
Digital Signal Processors (Cont.)
Functions
Voice termination
- Conversion between circuit-based voice and VoIP
Echo cancellation, VAD. jittermanagement
* Media TerminationPoint (MTP)
Passingone VoIP stream to another (same codec)
Transformation between a-lawand mu-law, or different
packetization periods
Transcoding
Conversionfrom one codec type to another
Conferencing
Mixing multiple voice streams
Ihe conversion between circuit- and VoIP-based voice is referred to as voice and media
termination. In addition tothis basic function, the DSPs offer other services. Ingeneral, the
functions of DSPsfall intothese four categories:
Voice termination: Applies toacall that has two eall legs, one leg on an analog ordigital
voice circuit and the second leg on a VoIP connection. The time-division multiplexing
(TDM) leg must beterminated by hardware that performs coding/decoding and
packetization of the stream. DSPs perform this termination function. Ihe DSPalso
provides echo cancellation. VAD. and jitter management at the same time that it performs
voice termination.
Media Termination Point (MTP): Bridges two full-duplex voice streams using thesame
codec, allowing the media streams to beset up and lorn down independently. TheMTP can
be used totranscode a-lavv to mu-law andviceversa, or it canbeused to bridge two
connections thai utilize different packeti/alion periods (different packet sizes).
Transcoding: Conversion from one codec toanother. Atranscoder compresses and
decompresses voicestreams to matchthe codecrequirements on bothcall sides.
Conferencing: Aconference bridge isa DSP that joins multiple participants into asingle
call. It can accept any number of connections for a given conference, uptothe maximum
number of streams allowed for a single conference on that device. There is a
correspondence between media streams that areconnected toa conference and participants
that are connected tothe conference. The conference bridge mixes the streams together and
creates a unique output stream for each connected party. Ihe output stream for a given
partv i-, the composite of the streams from all connected parties minus their own input
stream. Some conference bridges mix onlv the three loudest talkers on the conference and
distribute that composite stream toeach participant (minus their own input stream if they
are one of the talkers).
Digital Signal Processors (Cont.)
Branch
Central site
IVR
Using DSPs
The figure shows a multisite environment with deployed DSP resources.
The voice gateway in the central site offers transcoding and conferencing services. The
transcoding resources can beused totranscode G.729 toG.711 and then connect toan
interactive voice response (IVR) server.
The edge gateway in abranch isoffering DSP-based conferencing services to support mixed
codec environments and optimal WAN usage. The branch gateway with the analog telephone
attached toitperforms voice termination by converting an analog voice channel into VoIP.
)2010 Cisco Systems, Inc.
DSP Chip
Digital Signal
DSP Chip
* The DSP chip performs the sampling, quantization, encoding,
and opfional compression.
- The numberof simultaneous calls that a chipcan handle
depends on the type of DSPand the codec being used.
KE3
The DSP chip plavs a crucial role inthe Cisco Unified Communications system. The DSP chip
comes in several fonn factors, from soldered on to the main board of the Cisco Unified IP
phone or gateway, lo the modularpacketvoice DSPmodule (PVDM). The PVDM can have
multiple DSPs on the module.
fhe typeof DSPchip, the numberof DSPresources, and the type of codecthat is usedall
factor into thecalculation of howmany simultaneous callscanbe processed.
1-160 Implemerting Cisco Voice Communications and QoS (CVOICE) v8 0 ) 2010 Cisco Systems, Inc.
DSP Modules
Thissubtopic explains thetypes of high-density PVDMs.
DSP Modules
PVDM2 and PVDM3 Overview
PVDM2 installed in:
Motherboard PVDM2 slot on Cisco28CO and 3800 Series
ISRs
Cisco high-densitydigital voice network modules (NM-
HDV2, NM-HDV2-1T1/E1 and NM-HDV2-2T1/E1)
PVDM2 Adapter for PVDM3 slot on Cisco 2900, 3900
Series ISRs
PVDM3 installed in:
Motherboard PVDM3 slot on Cisco 2900, 3900 Series ISR
routers
- Cisco 2901 and 2911 routers haw 2 slots each, Cisco
2921 and 2951 routers have 3 slots each, and Cisco
3925 and 3945 routers have 4 slots each.
Cisco IOS Software Release 15.0.1(M) and later
Onboard 11 ftan.
Currently, there aretwomajortypes of high-density PVDMs: PVDM generation 2 (PVDM2)
and PVDM generation 3(PVDM3). The Cisco 2800 and 3800 Scries platforms support only the
PVDM2 modules. The Cisco2900and 3900 Series platforms support boththe PVDM2 and
PVDM3 modules. The PVDM3 modules provide higher density (up to four times higher) than
the PVDM2s. Theyalso provideimproved performance interms of the numberof conference
and transcoding sessions supported.
PVDM2 modules can be installeddirectly in the motherboard PVDMslot on Cisco 2800 and
3800 Series Integrated Services Routers. They canalsobe installed inthemotherboard PVDM
slots of the Cisco2900and 3900 SeriesIntegrated Services Routers usingspecial adaptercards
(PVDM2-ADPTR). Another option of using the PVDM2 on Cisco 2900and3900 Series
routers is to insert an NM-HDV2 module with PVDM2 modules into the service-module slot of
the Cisco2900and 3900Seriesplatforms usingthe network-to-server module adaptercard
(SM-NM-ADPTR).
The PVDM3 modules can be installed in the DSP slots on the motherboard of the Cisco 2900
and 3900 SeriesplatformsstartingwithCisco IOSSoftware Release 15.0.1(M). The PVDM3
modules cannot be installeddirectly on the PVDMslots of an NM-HDV2, only the PVDM2s
aresupported by theNM-HDV2. However, anNM-HDV2 that hasno PVDMs installed at all
can share PVDM3 DSP resources from the router motherboard PVDM slots across the chassis
backplane.
DSP Module Comparison
fhis subtopic prov ides a comparison between the PVDM2and PVDM3 Series modules.
PVDM2 / PVDM3 Comparison
Platform Cisco 2800, 3800, 2900,
support 3900 Series ISRs
Models PVDM2-8, PVDM2-16,
PVDM2-32, PVDM2-48,
PVDM2-64*
Capabilities Voice/Fax Voice/Video(NoCisco FaxRelay)
Resource Per-moduleand per-chassis DSP resources in motherboard slots
sharing sharing shared across the chassis backplane
Coexistence Can coewst on the Cisco 2900 and 3900 Series ISR platformsbut
PVDM2 cannot be installed directly on the motherboard
Cisco 2900,3900 Series ISRs
PVDM3-16, PVDM3-32, PVDM3-64,
PVDM3-128, PVDM3-192, PVDM3-256*
Njncer n ihe model name identifies the number of supported G 711 channels
The table in the figure lists the major differences between PVDM2 and PVDM3 modules. F,aeh
series includes multiple models that differ in ihe numberand capacity of the DSPsthat thev
have on board, fhe number(8. 16. 32. 64, and soon) in the model name indicates the
maximum numberof G.711 voice calls that a particularmodule can support.
All features supported bv PVDM2s aresupported on PVDM3s, except forCiscoFax Relay.
Cisco Fax Relav is no longer supported on PVDM3s. PVDM3 modules have a number of new
features, including video support.
fhe PVDM2 and PVDM3 modules can coexist as longas they are not both installed in the
same domain. The motherboard PVDM slots fonn one domain and each service module slot
forms a separate domain, 'fhe motherboard domain can contain either all PVDM2 modules or
all PVDM3 modules, Aserv icemodule domain canonly contain PVDM2 modules housed by
the NM-HDV2 carrier card. If a mix of PVDM2s and PVDM3s are delected on the
motherboardslots, then the PVDM2s will be deactivated, allowing only the PVDM3s to be
used activelv. If PVDM2s are detected in service-module slots and PVDMSs are installed on
the motherboard, then both will continue to function in their own domains and coexist.
1-162 Implementing Cisco Voice Communications and QoS (CVOICE) v8 0 2010 Cisco Systems. Inc
Codec Complexity
This topic describes codec complexity as it relates to DSP operations.
Codec Complexity
Media Termination and Transcoding
mWMmmmm
Lowcomplexity
Mediumcomplexity
High complexity
PVDM2-8
^^
(^
4
6
PVDM2-18 16
8
PVDM2-32
32
16
12
PVDM2-48
48
24
PVDM2-64
64
32
24
PVDM3-ie 16
12
10
PVDM3-32 32
21
PVDM3-64 64
42
28
PVDM3-128 128
96
60
PVDM3-192 192
138
68
PVDM3-256
256
192
120
Codec complexity refers to the amount of processing that is required to perform voice
compression. Codec complexity affects call density, which is the number ot calls that are
reconciled on the DSPs. With higher codec complexity, fewer calls can be processed. Ahigher
codec complexity may be required to support aparticular codec or combination of codecs. A
lower codec complexity supports the greatest number ofvoice channels, ifthe lower
complexity is compatible with the particular codecs in use.
<2010 Cisco Systems. Inc.
Codec Complexity (Cont.)
Recommended Usage in Deployment Models
Deployment model ' Recommended
codec
Single-sitedeployment -G.711/G.722
Multisite WAN with
centralized cal
processing, distributed
call processing, and
clustering over the IP
WAN
G.711/G.722for
tntrasite calls
G.729/A/B fa
int ersite calls
Oesciipticm
Codec used for VoIPcalls withinthe
same site where enough bandwidth is
available.
Intrasite calls consume more bandwidth
and provide the best voicequality.
fntersite calls consume littlebandwidth
and providegood voice quality.
RecommendedUsage in Deployment Models
This table e\plains the recommended codec choice in the various gateway deplovment models
Jhe selection of the appropriate codec depends on the VoIP path that the call takes as follows:
Single-site deplovment: In this deployment model, the VoIP calls are made within the
same sue. The site consists ofI.AN orMAN networks where enough bandwidth is
av ailable. G.7I I and G.722 codecs arc recommended lo provide the best voice quality, fhe
bandw idth usage ofthe codec is not aconcern within asingle site.
Multisite WAN with centralized or distributed call signaling and clustering over the
WAN: In these models, inlrasite calls should use the same codecs as in single siteG.711
or G.722- as these codecs oiler the best voice qualitv and the bandwidth consumption is
not aproblem, Inlersite calls should use (i.729 using any annex type. This codec family
consumes verv little bandwidth per call and guarantees good voice qualitv. It is widely'
supported in the industry, so the inleroperahility with other vendors is grained.
1-164
Implementing Cisco Voice Communications and QoS (CVOICE) v8.0
Packet Voice DSP Module Conferencing
This subtopic explains the DSP operation as aconferencing resource.
Packet Voice DSP Module Conferencing
Conferencing
* PVDM3 allows sharing the same DSP between transcoding, voice
termination, and conferencing.
PVDM2 allows sharing Ihe same DSP forvoice termination and
transcoding, buta dedicated DSP isneeded for conferencing.
The numberof supported conferencesand participants depends on
codec complexity. Forexample, PVDM3-256 supports:
66 G,711 conferences with8 participants each
- 6 G.711 conferences with 64 participants each
- 30 G.722 conferenceswith 8 participants each
- 36 G.729/G.729Aconferenceswith 8 participants each
- 18 iLBCconferences with 8 participants each
- Up to 32 participants per G.729/G.729A/G.722 conference
Upto 16 participantsper iLBC conference
PVDM2 modules offer more flexibility in resource sharing than the PVDM2 modules. The
PVDM3 modules have a universal firmware image that allows sharing DSP resources between
transcoding, voice, and conference calls. On the PVDM2, you can use the same DSP for voice
and transcoding calls, but adifferent DSP firmware image is required for conference calls. Ifa
PVDM2 DSP is assigned for aconferencing session, itcannot be used for transcoding or voice
callsat diesame time.Notethat conferencing needs a dedicated PVDM2 DSP. but not a
dedicated PVDM2 module. For example, the PVDM2-64 contains 4 DSPs; ifyou use one of
them for conferencing, theotherthree canbeused forotherpurposes.
The number ofsupported conferences and participants depends on codec complexity. As an
example, the PVDM3-256 module supports the following:
66 G.711 conferences with 8 participants each
36 G.729 or G.729Aconferences with 8 participants each
18 iLBC conferences with 8 participants each
Up to32 participants per G.729, G.729A, orG.722 conference
Up to 16participants per iLBC conference
DSP Calculator
This lopie presents anonline tool thatCisco ofTers for PVDM2 calculalion.
Calculate!
maom
DSP Calculator
Solstf 3ouTBi and Software Vn
SBSilKll
''|07Sio
Select ihe Router
Model
Select the Cisco IOS
Software Release |, uil
Cisco com http .'*mv Cisco conWcgi-DirVSupportiDSP'asp-calc.pi
There is an online tool that helps to calculate ihe PVDM2 resources that are needed for a
configuration. The DSP Calculator is available at: hUp:''A\\\\v,ci^co.coiii/Vgi-
bin.Support DSp'Jip-ude.p!.
Note The tool does not presentlyincludethe Cisco 2900 and 3900 Series routers and the PVDM3
modules
1-166 Implementing Cisco VoiceCommunications and QoS (CVOICE) v8 i
Configuring DSPs
This topic describes the configuration of DSP services as they relate to codec complexity.
Configuring DSPs for Voice Termination
Configurable Codec Complexity Options
High
- Support ahigh complexity codec or combination of high and lower
complexity codecs
Medium
.. Support amedium complexity codec or combination of med.um and low
complexity codecs
- Greatestnumberofvoicechannels
Flex
- Alows oversubscription-more w>ice channels can be^configured or
connected tothe module than the DSPs can accommodate
- If all voice channels goactiw simultaneously, some are unable lo
allocate a DSP resource
- Default setting
Secure
- Supports SRTP package capability
. Lowest number of selected medium-complexitycalls per DSP
The codec complexity can be configured lo tell the gateway how many DSP resources lo
allocate to avoice channel. These settings are available:
. High complexity: This option supports any high-complexily codec or acombination of
high- and lower-complexity codecs.
. Medium complexity: This option supports any medium-complexity codec or a
eombination of medium- and low-complexity codecs. It offers the greatest number of voice
channels, ifthe lower complexity is compatible with the particular codecs in use All
medium-complexity codecs can also be run in high-complexity mode, but fewer (usually
about half) ofthe channels arc available per DSP,
Flex: In this option, more voice channels can be connected (or configured in the case of
DSO groups and PRI groups) to the module than the DSPs can accommodate. If all voice
channels should go active simultaneously, the DSPs become oversubscribed, and calls that
are unable to allocate aDSP resource fail to connect. This is the default setting.
. Secure: This option supports the Secure Real-Time Transport Protocol (SRTP) package
capabilitv for media encryption and authentication. This setting supports the lowest number
of selected low- and medium-complexity codecs (G.711 a-law and mu-law, G.729. and
G.729A)perDSP.
Introduction to VoiceGateways 1-167
Configuring DSP Services for Voice Termination
fhis subtopic explains how to configure the DSP resources for voice
termination.
Configuring DSPs for Voice Terrninati
(Cont.)
Configuring Voice Card
router(eonflg)#
voice-card slot
Enters the voce cardconfiguration mode
router(conflg-volcecard]#
|dspfarm
Adds voice card toa DSP resource pool ~ " '
Local DSPs are available (or TDM streams on adifferent module or VWIC
router(conflg-volcecard)B
|codec complexity {flex : high ) medium | secur~
Specifies codeccomplexity
router(con fig-voicecard)W
codec sub-sample
Doubles Ihesampling frequency for G.711
he DSI resources uhen installed on avoice gateway, do not have to be configured to support
voice termination. In certain situations, it is necessary to fine-tune their operations For fine-
tuning, the voice-card slot command is used to enter the voice-card configuration mode The
voice card corresponds to aservice module installed on the gateway.
The dspfarni command adds aspecified voice card lo the DSP resource pool. If there are not
enough DSPs on the motherboard to terminate the required PRI and GAS ehannels, vou can use
hh PVDMî "rr n,pldetr f aVai'ab,e VOiW "^ (NM-' ")V2 r a',0ther nctwo* module
with PVDM2s) 1he DSPs of that voice card will be added to the shared resource pool This
method allows the termination of I'RI and CAS ehannels. but not analog circuits.
The codec complexity command sets the codec complexity on avoice card.
The codec sub-sample command is used for applications that have strict requirements for
round-lrip delay tunes. This command reduces the G.711 sampling period inside the DSP from
he default value o 10 ms to 5ms. thus reducing the delay. Ilovvever, this reduces the channel
density of 071 channels from 16 to 14. There is no difference in secure channel density ifthis
mode is enabled. J
Codec Complexity Configuration
This subtopic showsthe usageof the codec complexity and codec sub-sample commands.
Codec Complexity Configuration
Voice Card Configuration Example
router{config}# voice-card 1
router(config-voicecard)# codec complexity ?
flex Set codec Flex complexity, higher call density.
high Set codec to high complexity, lower call density.
medium Set codec to mid range complexity and call density
secure Set codec complexity to secure,
router(config-voicecard)# codec complexity flex
routertconfig-voicecard}# codec sub-sample
For codec complexity to change, all of the DSP voice channels must be in the idle state.
When you use the codec complexity high command to change codec complexity, the system
prompts you to remove all existing DSO or PRI groups using the specified voice card. Then all
DSPs are reset, loaded with the specified firmware image, and released.
i 2010 Cisco Systems. Inc. Introduction lo Voice Gateways 1-169
Configuring DSP Resources for Transcoding, Conferencing,
and MTP
This subtopic explains how to configure DSP resources for transcoding, conferencing, and
MTP.
ms
Configuring DSP
Transcoding, Confe
DSP-Farm Configuration
router(config)#
voice-card slot
Enters the voice card configuration mode
router(conflg-volcecard)B
dap services dspfarm
Enables DSP farm services
Makes the DSP services available for conferencing and transcoding
router(configl#
Idspfarm profile profile - '.6e.nl ifier {conference|mtp| transcode}
Creates a DSP farm profile for conferencing, MTP. or transcoding
To configure the DSP resources for transcoding, conferencing, or MTP, the DSP farms and
DSP farm profiles are used.
The dsp services dspfarm command creates a DSP farm and is available in the voice-card
configurationmode. A DSP farmis the collectionof DSP resources that are available for
conferencing, transcoding, and MI P services. DSP farms are configured on the voice gateway
and managed bv Cisco Unified Communications Manager through Skinny Client Control
Protocol (SCCP). The DSP farm can support a combination of transcoding sessions. MTP
sessions, and conferences simultaneously. The DSP farm maintains the DSP resource details
locally. Cisco [ 'nitled Communications Manager requests conferencing or transcoding services
from the gatevvav. which either grants or denies these requests, depending on resource
availabililv. Ihe details of whether DSP resources are used, and which DSP resources are used,
are transparent to Cisco Unified Communications Manager.
After the DSP serv ices have been made available, the dspfarm profile command is used to set
the required parameters. The DSP farmprofile indicates if the resource will he used for
conferencing, transcoding, or MTP,
1-170 Implementing Cisco Voice Communications and QoS (CVOICE) vB.O
Configuring DSP Resources for
Transcoding, Conferencing, and
DSP Farm Profile Configuration
router Iconfig-dspfarm-profile)W
I codec codec-type
Specifies the allowed codecs
router tconflg-depfarm-proflie)#
maximum sessions number
Defines the maximum number of sessions
router Iconfig-dspfarm-proflie)tt
I associate application seep
Enables SCCP for the profile
J
"H
DSP Farm Profile Configuration
DSP farm profiles are created toallocate DSP farm resources. Under the profile, you select the
service type (conference, transcode, orMTP). associate an application, and specify service-
specific parameters such as codecs and the maximum number ofsessions. ADSP farm profile
allows you to group DSP resources based on the service type. Applications associated with the
profile, such as SCCP, can use the resources allocated under the profile. You can configure
multiple profiles for the same service, each ofwhich can register with one Cisco Unified
Communications Manager group. The profile ID and service type uniquely identify a profile,
allowing the profile touniquely map toa Cisco Unified Communications Manager group that
contains a single pool of CiscoUnifiedCommunications Managerservers.
Transcoding and Conferencing Example
This subtopic prov ides an example ofhow to configure DSP resources for transcoding and
eonferencimi.
Transcoding and Conferencing
Example
Configuration Example
Cisco Uiified
Communications
Manager
10 1 1201 ~Cffi
dfipfacn profile 1 tran scode
codec g711ulaw
codec g711alaw
codec g729ar8
codec g725abr8
codec g72Sr8
maximum sessiong 6
associate application SCCP
no shutdown
dspfar i profile 1 conference
codec gTllulaw
codec gTlialaw
codec g729ar8
codec 9729abr8
codec g729rB
codec 9729br8
maxim im sessior s 2
aseoc ate appl cation SCCP
no shutdown
The figure illustrates the configuration oftwo DSP farm profiles. One isused for transcoding,
the other for conferencing. This example does notinclude the SCCP-rclated settings that are
required tomake the resources available tothe Cisco Unified Communications Manager, which
is out of the scope of this material.
Implementing Cisco Voice Communications and QoS (CVOICE] v8.0
>2010Cisco Systems, Inc
Verifying DSPs
This topic describes how toverify theDSP resources.
Verifying DSPs
routarl show voice
DSP DSP
TYPE NUN CH CODEC
COUNT
DSPMARE CURR BOOT P*K TX/RX
VERSION STATE STATE RST Rl VOICBPORT TS ABORT PACK
ap 0001 01 g711ula 0.1 busy 50/0/1,1
adap 0002 02 g729rB p 0.1 IDLE 50/0/1.2
edap 0003 01 g711ulm 0.1 IDLE SO/0/2.1
eosp 000* 02 g729r8 p 0.1 IDLE 50/0/2.2
___ __ FLBK VOICE CARD 0
DSP VOICE CHANNELS*
CURB STATE : (buByllnuae Ib-outibuay out (bpand) busyou t ponding
LEGEND : (bod)bad Ibhut)shutdc
[dpandldoimload pending
DSP DSP DSPMARE CURR BOOT
TYPE ND* CH CODBC VERSION STATE STATE RST AI VOICEPORT IS ABRT PACK COUHT
PAK TIt/BX
C551Q 001 01 gVllulaw as.6.0 buy idl . o 0 o/o/osis os
DSP SIGNALING CHANNELS*
0 228/9228
DSP DSPMARE CURR BOOT
PAK TK/RX
TYPE NUK CH CODEC
VERSION STATE 3TATS RST AI VOICEPORT TS ABRT PACK COUNT
-END OF FLEX VOICE CARD 0
The show \oicedsp command isused toverify codec complexity configuration. In this
example, avoice channel is served by aDSP with number 0001. The media channel uses G.711
mu-lawcodec and goes over the voiceport 0/0/0:15.
This topic explains how to verify the status of the DSP farms.
Verifying DSPs (Cont,
Verifying DSP Farm
Routers she- dspfarm dsp all
SLOT DSP VERSION STATUS CUNL USE TYPE
RSC ID BRIDGE ID PKTS TXED PKTS RXED
1 S 1 .0. UP H/A FREE conf
1
0 S 1.0.6 UP N/A FREE conf 1
Total number of DSPPARH DSP channel {si 2 -
P*Mi"" 1
H't.i 4.M-I...1I1 i '
To check ihe DSP status that was used for DSP farm proliles. use the show dspfarm dsp all
command, fhis example shows two available DSPs that have been configured for
conferen cina.
Implementing Cisco VoiceCommunications and QoS (CVO(CE) v8 0
)2010 Cisco Syslems, Inc.
Summary
Summary
Codecs compress and decompress data whenconverting
voice signals to VoIPpackets.
The MOS scores of major codecs representtheirperceived
quality and range between 3.5 and 4.3.
Total bandwidth of one VoIP conversation depends on the
packet ratio, codec, and Layer 2-to-Layer 4 overhead.
DSPs are specialized voice processors used for voice
termination, conferencing, transcoding, and MTP.
The number of voice channels served by one PVDM depends
on the codec complexity: low, medium, or high.
>Media resources available for voice termination can be fine-
tuned for codec complexity.
The status of media resources can be verified on the voice
gateway.
References
For additional information, refer to these resources:
PVDM2 DSP Calculator:
http:/wvvvv.cisco.coin'Cgi-bin/Support/DSP/'dsp-calc.pi
1-176 Implementing CiscoVoice Communications and QoS (CVOICE) w8 0 2010CiscoSystems. Inc
Module Summary
This topic summarizes the key points that were discussed in this module.
Module Summary
Voice gateways support theCisco Unified Communications
architecture byconverting voice signals andoffering
advanced voice features.
Call routing Involves incoming and outgoing call legs that
correspondto inbound and outbound dial peers.
Gateways support various interface types: analog with in-
band signaling (FXO, FXS, FXS-DID, E&M), digital with CAS
signaling (T1/E1 CAS), and digital with CCS signaling (T1/E1
PRI, BRI).
Voice conversion intoVoIP uses codecs with varying
complexify and MOS, and isperformed by dedicated DSPs.
This module describes the various types of voice gateways and explains how todeploy them in
different Cisco Unified Communications environments. It explains the eall routing process, the
direct inward dialing (DID) feature, and the various types ofvoice ports and their
characteristics, codecs, digital signal processors (DSPs), and their implementation.
References
For additional information, refer to these resources:
Cisco2900SeriesIntegrated Services Routers:
http:',<,vvvv\\ .cisco.com/go/2900
hup:'''uvvvv. cisco.com/go/3900
http'./.'vv vvvv.cisco.com/go/2800
http:'.www.cisco.coni/go/3800
Cisco ATA 186:
http:iVvvvvw.cisco.com/go/atal86
Cisco AS5350 Universal Gateway:
http:. \vvwv.cisco.com/go/as5350
Cisco AS5400 Series Universal Gateway platforms:
http'./.'www.cisco.com/go/iis5400
Cisco 7200 Series Routers:
http: w\\w.cisa>.eom.'go/7200
PVDM2 DSP Calculator:
http: www.cisco.com'ctii-b;n Support'DSP/dsp-calcpI
1-178 Implementing Cisco Voice Communications and QoS (CVOICE) vB.Q
Module Self-Check
Use the questions here to review what you learned in this module. The correct answers and
solutions are found inthe Module Self-Check Answer Key.
Q1) What differentiates Cisco Unified Communications from traditional telephony?
(Source: Understanding Cisco Unified Communications Networks and the Role of
Gateways)
A) the need for signaling protocols
B) bearercontrol
C) billingservices
D) messaging capabilities
Q2) Which two VoIP signaling protocols does aCisco Unified Communications gateway
support? (Choose two.) (Source: Understanding Cisco Unified Communications
Networks and the Role of Gateways)
A) RTP
B) SIP
C) SS7
D) MGCP
E) ISDN
Q3) Which two functionalities differentiate multisite WAN deployment with centralized
call processing from multisite deployment with distributed call processing? (Choose
two.) (Source: Understanding Cisco Unified Communications Networks and the Role
of Gateways)
A) intersite VoIPsignaling
B) codecsthat shouldbe used inthe WAN
C) PSTN signaling protocol
D) SRST
E) the need for DSPresources
Q4) Which feature is offered by the Cisco 2900 Series Integrated Services Routers?
(Source: Understanding Cisco Unified Communications Networks and the Role of
Gateways)
A) transmission speedsup to Tl/El
B) highest port density
C) PoE
D) 48 fully featured analog ports
Q5) ACisco Unified Border Element is agateway with at least one VoIP call leg. (Source:
Understanding Cisco Unified Communications Networks and the Role ofGateways)
A) true
B) false
06) How manv inbound call legs arc associated with anormal end-to-end call that goes
over an IP WAN'? (Source: Examining Gateway Call Routing and Call Legs)
A) one
B) two
C) three
D) four
Q7) Which dial peers should be configured lo complete an end-to-end call? (Source:
Examining Galewav Call Routing and Call Iegs)
A) the inbounddial peers onlv
B) the outbound dial peers onlv
C) one inbound dial peer and one outbound dial peer
D) all four dial peers
08) Arrange the steps in the call-setup process in the correct order. (Source: Examining
Gatewav Call Routing and Call Legs)
I 1'he terminating gatewav creates the inbound voice network call leg and
assigns it a call ID.
2, fhe POTS call arrives at the originating gateway, and an inbound POI S
dial peer is matched.
-' 'he originating gatewav creates an outbound voice network call leg and
assigns it a eall ID.
4- Ihe originating gateway creates an inbound POTS call leg and assigns it a
call ID.
. -s- J'he uiice network call request arrives at the terminating gateway, and an
inbound voicenetwork dial peer is matched.
- - '1'he terminating galewav creates an outbound POTS call leg and assigns it
a call ID.
1- 'llt: terminating gatewav uses the dialed string to match anoutbound
POTS dial peer.
._ 8- At this point, both gatewavs negotiate voice network capabilities and
applications, if required.
y- ' heoriginating galewav uses thedialed string (omalch anoutbound voice
network dial peer.
Q9) Whal is the role ofthe terminating router ifthe originating router requests nondefault
voice capabilities? (Source: Examining Gateway Call Routing and Call Legs)
A) Ihe terminating router negotiates with theoriginating router until the default
capabilities are accepted.
Rl The terminating router reconfigures the ports tomeet the requested capabilities.
C) Iheterminating router matches an inbound voice network dial peer that has the
requested capabilities.
D) The terminatingrouter terminates the call.
1-180 Implementing Cisco Voice Communications and QoS (CVOICE) v8 0 2010 Cisco Systems, Inc.
QIO) Which two functions are perfomied by aPOTS dial peer? (Choose two.) (Source:
Examining Gateway Call Routing and Call Legs)
A) providing an address for the edge network or device
B) providing adestination address for the edge device that is located across the
network
C) routing the call across the network
D) identifying the specific voice port that connects the edge network or device
K) associating the destination address with the next-hop router or destination
router, depending on the technology that isused
Q11) What is the term for the address is that is configured on the dial peer? (Source:
A) telephone number
B) number range
C) destination pattern
D) call endpoint
Q12) Which two parameters must be specified on arouter that is connected to atelephone?
(Choose two.) (Source: Examining Gateway Call Routing and Call Legs)
A) voice port
B) dial type
C) calling plan
D) telephone number
E) dsO-group
Q13) Which router or routers require the configuration ofaPOTS dial peer? (Source:
A) one inbound router on thenetwork
B) oneoutbound router onthe network
C) one inbound and one outbound router on the network
D) each router where edge telephony devices connect to the network
Q14) What does aplus (+) sign before the telephone number indicate? (Source: Examining
Gateway Call Routing andCall Legs)
A) The telephone number must conform to ITU-T Recommendation E.164.
B) The number isanextension ofa telephone number.
C) An additional digit must be dialed before the telephone number.
D) Thetelephone number canvary inlength.
Q15) Which special character in adestination pattern string is used as awildcard? (Source:
A) asterisk (*)
B) pound sign (#)
C) comma (.)
D) period (.)
016) What happens when no matching dial peer is found lor an outbound call? (Source:
A) fhe defaultdial peer is used.
Bl Dial peerO is used.
C) Ihe POTS dial peer is used.
D) "fhe call is dropped.
QI7) Which parameter is configured only for POTS dial peers? (Source: Examining
Gateway Call Routing andCall Legs)
A) answer-address
B) destination-pattern
C) incoming called-number
D) port
Q18) hi which sequence does agalewav malch the called number in the call-setup request to
adial-peer attribute'.* (Source: Examining Gateway Call Routing and Call Legs)
_ answer-address
destination-pattern
port
019) Match the dialed number to its most specific dial peer. (Source: Examining Gatewav
Call Routing and Call Legs)
A) 5551234
B) 5553000
C) 5553216
D) 5554123
i. dial-peer voice 1 pots
destination-pattern .T
port 1/0:1
2. dial-peer voice 2 pots
destination-pattern 555|0-2,5|...
port 1/1/0
3. dial-peer voice 1 pots
port 1/0:1
_. 4. dial-peer voice 1 pots
port 1/0.1
020) The serv ice isoffered bv telephone companies, and it enables callers todial an
extension directly on a VoIP system. (Source: Examining Gateway Call Routing and
Call 1egs)
021) Which tvpe of signaling is used when a router provides current on the E-lead as soon
as it sees current onthe M-lead? (Source: Configuring Gateway Voice Ports)
A) delav-start signaling
B) immediate-start signaling
C) ground-star! signaling
D) OS1G
1-182 Implementing Cisco Voice Communications and OoS (CVOICE) v8.0 2010 Cisco Systems, Inc
Q22) Which configuration parameter sets the dial tone, busy tone, and ringback tone?
(Source: Configuring Gateway Voice Ports)
A) cptone
B) ring frequency
C) ring cadence
D) description
E) signal
Q23) ACAMA trunk is aspecial analog trunk type that was originally developed for long
distance billing but is now mainly used for^__. (Source: Configuring Gateway
Voice Ports)
Q24) Which parameter should be configured when you want to set alimit on the number of
seconds towait between dialed digits before digit input evaluation? (Source:
Configuring Gateway Voice Ports)
A) timeouts initial
B) timeouts interdigit
C) timing digit
D) timing interdigit
Q25) When E1 R2 is being used, time slot is used for signaling, and each of its frames
carries information for voice time slots. (Source: Configuring Gateway Voice
Ports)
A) 23. three
B) 23. two
C) 17. three
D) 17. two
Q26) The two types of ISDN interfaces arc and . (Source: Configuring Gateway-
Voice Ports)
A) Tl. El
B) PRI, BRI
C) TI.PRI
D) El. BRI
Q27) ISDN uses for Layer 2signaling, which is defined in _. (Source:
A) EAPB.Q.931
B) LAPD. Q.931
C) LAPB. Q.921
D) LAPD. Q921
Q28) The command is used to configure aTl controller for CAS. (Source:
Configuring Gateway VoicePorts)
A) pri-group
B) bri-group
C) dsO-group
D) dsl-group
029) Which condition must occur for echo to become aproblem? (Source: Configuring
Gatewav Voice Ports)
A) disabled echo canceller
B) sufficient voice amplitude
C) leakage between transmit (Tx) and receive (Rx) paths
D) incorrectly selected lie-line (two-wire versus four-wire)
Which command prov ides information about the echo canceller settings? (Source-
A) show voice call
B) show voice port
C) test voice controller
D) show controller
030)
03
033;
Arrange the codecs in descending order ofbandwidth requirements (from highest to
lowest bandwidth). (Source: Understanding DSP I-unctionality, Codecs, and Codec
Complexity )
_ . 1. il.BC
2, G.711
_ 3. G.726r32
4. G.729
5. G.723r53
032) Which statement describes G.729 Annex B? (Source: Understanding DSP
functionality. Codecs, and Codec Complexify)
A) It is interoperable with G.729 Annex A.
B) It uses lower bandwidth than (i.729 Annex A.
C) It is more susceptible to delay, variation. and "tandeming" than G.729 Annex
D) it has higher complexity than G.729Annex A.
Which ofthese is ascoring system for voice quality? (Source: Understanding DSP
Functionality. Codecs, and Codec Complexity)
A) PSQM
B) MOS
C) PESO
D) SCORE
034) Arrange the encapsulations in descending order ofoverhead (from highest lo lowest
overhead). (Source: Understanding DSP Eunclionality. Codecs, and Codec
Complexity)
Ethernet
PPP
MPLS
GRE
ESP
1-184 Implementing Cisco Voice Communications and QoS (CVOICE) vS 0 2010 Cisco Systems Inc
Q35) Which two statements describe PVDM2 and PVDM3? (Choose two.) (Source:
Understanding DSP Functionality. Codecs, and Codec Complexity)
A) Both can be installed on router motherboards.
B) Both can be installed in appropriate PVDM adapters.
C) Both support fax and voice.
D) Both support voice and video.
E) Both can be installed in aCisco 3900 Series Integrated Services Router
platform.
Q36) The highest number of voice calls that are supported on aDSP is achieved with the
codec complexity set to . _. The lowest number of voice calls that are supported on
aDSP is achieved with the codec complexity set to __. (Source: Understanding
DSP Functionality, Codecs, and Codec Complexity)
Q37) Which information does the show voice dsp command display? (Source:
Understanding DSP Functionality, Codecs, and Codec Complexity)
A) codecoverhead
B) packet statistics
C) average delay
D) average packet loss
QD
D
021 B. D
Q?>
A. [)
041 C
Q5) IS
061 B
07) I")
081 1 Ih.
Ilie POTS call armes at ihe oriym;iting gateway, and an inbound POTS dial peer is matched. (2.)
2 Hie originating gatewav creates an inbound POTS eall leg and assigns it acall ID (4 )
- 'I lie ontiinatinii galewav uses the dialed stung to match an outbound voice network dial peer. (9.)
4 Ihe originating gatewav creates an outbound voice network call leg and assigns it acall ID. (3 }
:> [iie voice network eall request arrives ai the terminating gateway, and an inbound voice network dial
peer is matched | s I
6. Ihe terminating gateway creates the inbound voice network call leg and assigns it acall ID. (1 )
' At tins point, both gatewavs negotiate voice network capabilities and applications, ifrequired. (8.)
8 Ihe terminating gateway ii^s the dialed siring to match an outbound POTS dial peer. (7.)
'> 1he terminating gateway cieates an outbound POTS call leg and assigns ilacall ID (6.)
09,
c
010) A. 1)
OH) C
Q12) A. D
QI.m D
Q14] A
t;i5) D
Q1 fi i D
OPi
D
018) 1 incoming called-number
2 answer-address
? dcslination-patiern
1 pari
Qll>) 1-D
2-.\
>-B
4-C
Q20) direct inwarddialing(DID)
0211 B
022) A
Q2l) eincrgenev call sen ices such a.-, 9| 1
0241 B
Q25) D
02'', B
0--)
D
028) C
02v) C
1-1
86 Implemenling Cisco Voice Communications and QoS (CVOICE) v8 0 2010 Cisco Systems Inc
Q301 B
031)
1 G ' 11
2. G.726r32
3 iLBC
4 G729
5 G 723r53
Q32) D
Q33) B
Q34) 1 ESP
2 GRF.
3. Fthernet
4. PPP
5. MPLS
035) A. II
Q36) medium, secure
Q37| B
Module 2
VoIP Call Legs
Overview
VoIP transmission differs from traditional circuit-switched telephony in the way that the calls
arc signaled and the voice media is transported through the network. Successful implementation
ofaVoIP network relies heavily on the correct deployment ofVoIP gateway signaling
protocols' H.323. Session Initiation Protocol (SIP), and Media Gateway Control Protocol
(MGCP) The VoIP network provides special transmission methods tor fax, modem, and dual
tone multifrequency (DTMF) tones. This module describes the characteristics and
implementation of the gateway signaling protocols, and explains the configuration of VoIP dial
peers, including advanced features such as fax/modem pass-through and relay, and DTMF
relay.
Module Objectives
Upon completing this module, you will be able to describe the characteristics and configuration
elements ofVoIP call legs. This ability includes being able to meet these objectives:
Describe how VoIP signaling and media transmission differs from traditional voice circuits
and explain how voice is sent over IP networks, including analog-to-digital conversion,
encoding, packetization. and all variants ofRTP
Describe the characteristics of H.323 and explain when louse it
Describe the characteristics of SIP, and explain when to use it
Describe diecharacteristics of MGCP and explain when louse it
Discuss special requirements for VoIP call legs, including the need for QoS. fax/modem
relay and DTMF support
Describe howto configure VoIPcall legs in a gateway
2-2 Implementing Cisco Voice Communications and QoS (CVOICE) v8.0 2010 Cisco Systems Inc
Lesson 11
Examining VoIP Call Legs and
VoIP Media Transmission
Overview
The inherent characteristics of aconverged voice and data IP network present certain
challenges to network engineers and administrators in delivering v01ce traffic correctly. Ihis
lesson describes the challenges of integrating avoice and data network and explains the
technologies that enable voice media transmission.
Objectives
Upon completing this lesson, you will be able to describe how VoIP signaling and media
transmission differ from traditional voice circuits. Further, you will be able to explain how
voice is sent over IP networks. You will also be able to describe analog-to-d.glial conversion,
codecs, voice activity detection (VAD), packetization, and all variants of Real-Time Transport
Protocol (RTP). This includes compressed Real-Time Transport Protocol (cRTP),!Secure Real-
Time Transport Protocol (SRTP). and Real-Time Transport Control Protocol (RTCP). This
ability includes being able to meet these objectives:
Describe how voice is transported over IP networks end-to-end, and compare traditional
versus VoIP signaling and transmission mechanisms
Describe the process of analog-to-digilal voice conversion using PCM, including optional
voice compression
. Describe the process of digital voice packetization and explain which options affect packet
size
Describe the characteristics of the four protocols used for media transmission in aVoIP
network and outline their limitations
Describe the process of suppressing silence to conserve per-call bandwidth, and list options
that exist to accomplish this process
VoIP Overview
This topic describes the key differences between VoIP and traditional telephonv and the mak
stages ot \oice processing in VoIP. " J
IP and Traditional Telei
Transmission
technology
Basic signaling
funclions
Signaling
protocols and
methods
Transmission
method
Traditional telephony
Circuil -switched
Supervisory, address, informational
Digital: SS7, ISDN, QSIG
Analog:loop-start, ground-start,
immediate-start, wink-start, delay-start
DTMF, pulse
Dedicated crcuit
Packet-switched
Supervisory, address,
informational
H.323, SIP, MGCP,
SCCP
Bundle of UDP flows
Voll transports \oiee information over IP networks, which use packet-switched forwarding
Ihis principle differs from the circuit-switched technology of traditional telephone networks
where achannel is set up between the communicating endpoints through the
telecommunications infrastructure.
Before acall is established, signaling methods are used to delect off-hook stale collect the
called number, and inform the network aboul the call. The signaling protocols fulfill similar
Junctions, and must meet additional requirements imposed by the IP-based transmission
method, for example, negotiation ofVoIP transmission parameters such as codecs.
There are four VoIP signaling protocols: 11.323. Session Initiation Protocol (SIP) Media
Cmtewaj Control Protocol (MGCP), and Skinny Client Control Protocol (SCCP). each ofthem
best suited for specific scenarios,
Ihe media is transported in IP networks in RTP packets that are encapsulated in User Datagram
Protocol (I. DP) flows. Asingle voice conversation involves several UDP flows.
2-4 Implementing Cisco Voice Communications and QoS (CVOICE) v8.0
Major Stages of Voice Processing in VoIP
This subtopic lists the major stages of how voice information is converted and transmitted in a
VoIP environment.
Major Stages of Voice Processing in VoIP
1. Sampling
2. Quantization
3. Encoding
4. Codec Compression
5. VoIP Encapsulation
7. VoIP Decapsulation
8. Decoding
9. Modulation
4%/
Voice Gateway j
6. TransportThrough IP Network
For transmission over an IP network, the voice wavelength must be sampled, quantized,
encoded, optionally compressed, and then encapsulated in aVoIP packet. The first four steps
are perfomied bv adigital signal processor (DSP) in the originating gateway and are detailed
later in this lopie The VoIP packets are then delivered to the destination gateway and the voice
infonnation is retrieved from the packet. Finally, aDSP on the terminating gateway decodes
the payload and modulates the wavelength to reverse the process performed on the onginating
gateway.
VoIP Call Legs
2-5
VoIP Components
This subtopic describes the components ofa VoIP environment.
VoIP Componen!
Application
Server
4i
Mu'tipo nt Control a.
Ihrt **
Cisco Unrfiea IP Phones
Videoconferenc
Station
Gatekeeper
n , IP Backbone
Ga'e^y Gateway pBX
Most VoIP devices terminate signaling and media
Cisco UBE processes signaling andoptionally
proxies media
' Gatekeeper provides Call Admission Control
(CAC) by responding to endpoint queries
Ihe figure depicts the basic components ofapacket voice network:
i Cisco I'nified IP phones: Provide an IP endpoint for voice communication.
i Gatekeeper: Provides Call Admission Control (CAC). bandwidth control and
management, and address translation.
i Gateway: Provides translation between VoIP and non-VoIP networks such as public
switched telephone network (PSTN). Gateways also provide physical access for local
analog and digital voice devices such as telephones, fax machines, key sets, and PHXs.
I Cisco I'nified Border Element: Interconnects two VoIP networks. It acts as aproxy
between signaling protocols and may optionally proxy the media stream.
Multipoint control unit: Provides real-time conncctiv ity for participants in multiple
locations to attend the same videoconference or meeting.
Call agent: Prov ides call control for Cisco Unified IP phones. CAC, bandwidth control and
management, and address translation.
Application servers: Provide services such as voice mail, unified messaging, interactive
voice response (IVR). presence information, multimedia conferencing, and others.
Videoconference station: Provides access for end-user participation in videoconferencing.
The vidcoconference station contains avideo capture device for video input and a
microphone for audio input. The user can view video streams and hear the audio that
originates at a remote user station.
2-6 Implementing Cisco Voice Communications and QoS(CVOICE! v80
This topic provides an overview of converting voice information to VoIP.
Overview
Pulse Code
Modulation
(PCM)
Codec
Compression
1. Sample theanalog signal regulariy.
2 Quantize the sample.
3 Encode the value into a binary expression.
4 Compress the samples to reduce bandwidth
(optional).
The table describes the steps toconvert voice information toVoIP:
Step Procedure
Description
1.
Sample the analog signal regularly.
Thesampling ratemust be twice the highest frequency
toproduce playback thatdoesnot appear either choppy
or too smooth. The samplingrate used intelephony is
8000samples persecond (8kHz), which reflects the
fact that the bulk of human voiceenergyis carried inthe
spectrum of 0-4 kHz.
2.
Quantize the sample.
Quantization consistsof a scale made up of 8 major
segments. Each segment is subdivided into 16
intervals. Trie segments are not equally spaced but are
actually finest near the origin. Intervals areequal within
the segments but different when they arecompared
between thesegments. Finer graduations at theorigin
result in less distortion for low-level tones.
3.
Encode the value into an 8-bit
digital form.
Encoding mapsa value derived from the quantization to
an 8-bit number (octet).
4 (Optional) Compress the samples
to reduce bandwidth
Signal compression is used toreduce thebandwidth
usage per call.
The first three steps describe the pulse code modulation (PCM) process, which corresponds to
the G.711 codec. Step 4explains compression that is performed by low-bandwidth codecs, such
asG.729. G.728. G.726. or Internet Low Bitrate Codec (iLBC).
VoIP Call Legs 2-7
Sampling
2-8
This subtopic explains the sampling process.
Sampiim
Significant human articulation range
300 Hz to 4 kHz
Nyquist theorem, sampling rate =2xmaximum articulation frequency
2x4 kHz = 8kHz = 8000/sec
Each sampleis 1/8000 ofa secondapart
Analog Waveform
A
\
/
J. Il\ Time
"1 f~
X f
\L F
Sampling is aprocess that takes readings ofthe waveform amplitude at regular intervals, bv a
process called pulse amplitude modulation (PAM). Ihe output is ascries ofpulses that
approximate the analog waveform. For this output to have an acceptable level ofqualitv for the
signal to be reconstructed, the sampling rate must he rapid enough.
Harry Nyquist developed amathematical proof about the rate at which a waveform can be
sampled and the infonnation that can be recovered from those samples. The Nyquist theorem
states that when asignal is instantaneously sampled at the transmitter in regula'r intervals and
has arate ofat least twice the highest channel frequency, the samples will contain sufficient
information to allow an accurate reconstruction ofthe signal at the receiver.
While the human ear can sense sounds from 20 to 20,000 IU. speech encompasses sounds from
about 200 to 9000 11/, 'fhe telephone channel was designed lo operate at frequencies of300 lo
4000 Hz. This economical range offers enough fidelity for voice communications, although
higher frequency tones are not transmitted. The removal ofhigher frequencies leads to issues
with sounds such as "s" or "th." The voice frequency of4000 11/ requires 8000 samples per
second: that is. one sampleevery 125 microseconds.
Implementing Cisco Voice Communications and QoS(CVOICE) v8.0
Quantization
Thissubtopic describes thequantization process.
Quantization
G.711 Operations
Segment2
Segment 1
Segment 0
Segment 0
Segment 1 Z
Segment 2
Each sample is 1/S000 of a
second apart
Types:
mu-law
a-law
Time
Quantization divides the range ofamplitude values that are present in an analog signal sample
inlo aset ofdiscrete steps that are closest in value tothe original analog signal. Each step is
assigned aunique digital codeword. Quantization matches a PAM signal to asegmented scale.
The scale measures the amplitude (height) ofthe PAM signal and assigns an integer number to
define that amplitude.
"ITie figure shows quantization. In the example, the x-axis represents time and the y-axis
represents the voltage value (PAM). The output isaseries ofpulses that approximates the
analog waveform.
The voltage range isdivided into 16 segments (0to7positive, and 0to7negative). Starting
with segment 0.each segment has less granular intervals than the previous segment, which
reduces the signal-to-noise ratio (SNR) and makes the segment uniform. This segmentation also
corresponds closely tothe logarithmic behavior of thehuman ear.
The two principal schemes for generating these samples inelectronic communication are a-law
and mu-law.
VoIP Call Legs 2-9
Quantization (Cont
G.711 a-law and mu-law
Similarttfes bMwmn mu-law and a-taw
Bothare linear approximations of logarithmic
input/output relationshp.
Both are implemented using eight-bit codewords
(256 levels, one for each quantization interval).
Bolh break trie range into a total of 16
segmenis:
Eight positive and aght negative segments.
Each segment is twicethe length of the
preceding one.
Uniformquantization in each segment.
Bothuse a siniar approach to coding the eight-
bit word:
First (MSB] identifies polarity
Bits two. three, and four identify segment
Final four bits quantize the segment are the
lower signal levels than a-law
Different linear approximations lead lo
different lengths and slopes.
The numerical assignment of the bit
positions in the eight-bit codeword to
segments and the quantization levels
within segments are different.
a-law provides a greater dynamic range
than u-law.
mu-law provides better signal/distortion
performancefor low-level signals than a-
law.
An international connection needs lo
use a-law
mu to a conversion is the res pons ibility
of Ihe mu-law country.
a-law and mu-law arc audio compression schemes defined by ITU-T G.711, that compress 16-
bit linear PCM data down toeight bits oflogarithmic data, a-law standard isprimarily used in
Europe and the rest of the world, mu-law is used inNorth America and Japan.
fhe similarities between mu-law and a-law include thefollowing:
Both arc linear approximations of logarithmic input/output relationship.
Both are implemented using eight-bit codewords (256 levels, onefor each quantization
interval). Fight-bit codewords allow for a bit rate of 64kb/s. This iscalculated by
multiplying thesampling rate(twicethe input frequency) bv the size of the codeword (2*4
kHz* 8 bits = 64 kb/s).
Both break adynamic rangeintoa total of 16segments:
fight positive and eight negative segments.
Kach segment is twice the length of the precedingone.
1!nifonn quantization is used within each segment.
Bothuse a similar approach to coding the eight-bit word:
first bit (MSB) identifies polarity,
Bits two. three, and four identify segment.
final four bits quantize the segment are the lowersignal levels than a-law.
Thedifferences between mu-law and a-lawinclude the following:
Different linearapproximations leadto different lengths and slopes.
The numerical assignment of the bit positions in theeight-bit codeword lo segments and the
quanti7ation lev els within segments are different.
a-law provides a greater dynamic range than mu-law.
mu-law provides better signal-distortion performance for low-level signals than a-law.
Implementing Cisco Voice Communications and QoS (CVOICE) v8 0 )2010 Cisco Systems. Inc
its for auniform PCM equivalent, mu-law requires 14 bits for auniform
-lawto a-law conversion is the
a-law requires 13bits
PCMequivalent.
An international connection must use a-law. mu
responsibility ofthe mu-law country.
VoIP Call Legs 2-11
Encoding
2-12
This subtopic describes the process ofencod
nig.
G 711 8-Bit Words
HHQHHHH
Sign:
1 = Positive
0 = Negative
~Y
Segment
V V-
Inferval
Example: mu-law = +99 and a-law=+28
I.coding com ens an integer base-l 0number to abinary number, fhe output of encoding is a
innu r,"01' m ' ,"? h" " elthCr 3' ,PMr a<no P"1"*- Ater ^M samples
an input analog vo.ee signal, the next step is to encode these samples in preparation for
transmission ov era telephony network, "fhis process is called PCM.
The PCM process mathematically converts the value obtained from PAM sampling to another
b, ry value- within the range -127 to +127. I. is at this stage that companding, the proccstof
first compressing an analog signal at the source and then expanding this signal back to its
original size when ,t reaches its destination, is applied. This whole process is generally referred
lo as PC Mcoding. ADSP. which is aspecialised chip, performs the PCM process quieklv
In the United States. Canada, and Japan, mu-law is used. The rest of the world uses a-law. Both
mu-law and a-law companding produces PCM values in the range of-127 to +127 Both mu-
aw and a-law represent apositive sign value with avalue of I. and anegative sign with avalue
Zny^:^T'^"^ 'r0m ,hC "nnnar rampuWi0naI - ^ "**
Of the two methods, a-law appears to be the more logical method, because aPCM value of
-127 is represented as 11111111. In other words, apositive sign value (Ihe first bit) followed
bv abinary valueof 127 composed ofthe segment and interval bils. Similarly -12 is
represented as 00100000. Mu-law operates ahi, differently by logically inverting .he segment
and interv al bits. ,s,ng mu-companding. ihe value of t 127 becomes 10000000, in other words
a positive sign value (the first bit) followed by ihe bit inverse of+127.
Note
When amu-law connects with an a-law country, the mu-law end must convert the signal.
Implemenling Cisco Voice Communications and QoS (CVOICE) vB.O
Compression
"fhis subtopic explains compression.
Compression
Optional
iîîîV ! Bandwidth fkh/sl îî^H
G.711 64
G.726KJ2 32
G.726r24 24
G.726M6 16
G.728 16
iLBC 15.2.13.3
GSM Full Rate (GSM-FR) 13
G.729 (A/B/AB) 8
G.723r63 6,3
G.723r53 5.3
"iLBC = Internet Lew Bitrate Codec
Uncompressed digital speech signals are sampled at a rate of 8000 samples per second, with
each sample consisting of 8 bits. This corresponds to 64 kb/s per call. Multiple algorithms have
been developed to allow voice transmission at lower bandwidth consumption. The most
common codcr-dccodcr (codec) algorithms are presented in the table in the Figure,together
with their bandwidth.
i 2010 Cisco Systems, Inc. VoIP Call Legs
VoIP Packetization
fhis topic describes tiie process of packetization.
PCM (G.711)
10010111 Sample!
10010110 Sample 2
10010101 Sam pie 3
10010100 Sample 4
10010011 Sample5
10110001
10010111 10010110 10010101 10010100 10010011 1 . | 10110001 |
G 711 20 ms of samples (160 bytes)
G 711 30 ms of samples (240 bytes)
Alter the voice wavelength is digitized, the DSI* collects the digitized data for an amount of
time until there is enough data to fill the payload of a single packet.
"flicexample in the ligure shows how I'CM samples are packaged into the payload of a single
packet using G.71 1codec. With G.71 I. either 20 ms or 30 ms worth of voice wavelength are
transmitted in a single packet.
20 ms worth of \ oicc wa\ elcngth corresponds to 160 samples (there arc 8000 samples per
second. 10 ms would correspond to 80 samples and 20 ms would be 160 samples). With 20 ms
worth of voice wavelength. 50 VoIP packets are transmitted in each direction in one second (1
second consists of 50 20-ms intervals: I see / 20 ms - 50).
SimilarK. 30 ms worth of voice wavelength corresponds to 240 samples (there are 8000
samples per second. 10ms would equal 80 samples and 30 ms would be 240 samples). With 30
ms worth of voice wavelength, approximately 33 VoIP packets arc transmitted in each direction
in one second (1 second consists of 33.|3J 30-ms intervals: I sec / 30 ms -= 33.[3]).
2-14 Implemeiting Cisco Voice Communications and QoS (CVOICE) v8 0
Packetization Rate
This subtopic lists the common packetization rates and the resulting bandwidth figures.
Packetization Rate
The length of voice streams in a packet affects packetizatior
sample size, and voice bandwidth.
rate,
20-ms
voice
length in
a packet
30-ms
voice
length in
a packet
40-ms
voice
length in
a packet
60-ms
voice
length in
a packet
80-ms
voice
length in
a packet
Packetization rate 50pfe 33.3 p/S 25 p/s IBJp/s 12.5 p/s
Size of colected
G.711 samples for
a single packet
160 B 240 B 320 B 480 B 640 B
Uncompressed raw
voce bandwidth
64 kb/s 64 kb/s G4kb/3 64 kb/s 64 kb/s
Layer 3+
uncompressed
VoIP bandwidth
80 kb/s 74.7 kb/s 72 kb/s 69.3 kb/s 68 kb/s
The length of voice information carried in a single packet affects the payload size, which is
referred to in the table as the size of collected G.711 samples for a single packet. Before the
payloads arc transmitted over the IP network, they must be encapsulated in a packet that
introduces an additional overhead caused by Open Systems Interconnection (OSI) Layers 3 and
above. These headers consume additional bandwidth, in addition to the 64 kb/s required for raw
voice transmission. The bandwidth overhead depends on packet rate, as shown in the table.
i 2010 Cisco Systems, Inc. VoIP Call Legs 2-15
Codec Operations
2-16
This subtopic explains how codecs aredeployed to limit Ihe bandwidth consumption.
Codec Operations
G.729
DSP
* *" *^Hii
^^Tp^^'H
I
1^1 . | 1mi 1 n. j nm- J
| Codewsrd || Ctoema || Codewird || Codewo 11 Codeward 11 Codeword |
1 \\ *l 1
1 1 t
Payload Payload Payload |
By default one packet contains 20 ms of voice. 2 codewords
30 ms packetization period. 3 codewords in one packet
This ligure illuslrates the operation of an optional codec algorithm. G.729 is presented in this
example, fhe DSPsamples, quantizes, and encodes the analog wavefonii at the input. The DSP
generates one codeword for each 10 milliseconds worth of voice. The codewords are
encapsulated in the pavload of VoIP packets. A single VoIP packet carries by default 20 ms of
audio, encapsulating two G.729 codewords in one payload. Another supported packetization
rate is 30 ms. in which the VoIP packets are generated every 30 ms and carry three G.729
codewords in each packet.
Implemeitmg Cisco Voice Communications and QoS (CVOICE) v8 0 2010 Cisco Systems, Inc.
Packetization and Compression-G.729 Example
This subtopic presents the relevant parameters that describe the operation options of the G.729
codec.
Packetization and Compression Example
G.729
Layer 3 +bandwidth per call =(Voice Payload +Layer 3
Overhead[40B]) x Packets per Second x 8 bits/Byte
20 ms 30ms
voice length
In a packet
voice length
in a packet
Packetization rate
50 p/s 33.3 p/s
Size of collected, compressed
G.729 samples tor a single packet
20 B 30B
Compressed raw voice bandwidth
8 kb/s 8 kb/s
Layer 3+ G.729 VoIPbandwidth
24 kb/s 18.7 kb/s
This figure illustrates the common operation modes ofthe G.729 codec: 50-p/s rate with 20 ms
ofvoice wavelength in asingle packet, and 33.3-p/s rate with 30 ms of voice wavelength in a
single packet. After compression, the payload size is 20 bytes, or30 bytes, respectively. In both
modes, the compressed raw voice bandwiddi is8kb/s but the Layer 3+ bandwidth depends on
the packetization rate, and is 24 kb/s and 18.7 kb/s, respectively.
The call bandwidth can be computed using the formula:
Bandwidth per Call =(Voice Payload +Layer 3Overhead +Layer 2Overhead) *Packets per
Second * 8 bits/Byte
The examples shown in the table do not consider Layer 2overhead, which varies based on the
packet technology in use.
VoIP Call Legs 2-17
VoIP Media Transmission
Thistopicdescribes the protocols used for mediatransmission in the IPnetwork.
VoIP iViedia Transmission Overview
Real-Time Transport Protocol: Delivers the actual audio and video
streams over networks.
Real-Time Transport Control Protocol: Provides out-of-band
control information for an RTP flow.
cRTP compresses IP/UDP/RTP headers on low-speed serial links.
SRTPprovides encryption, message authentication and integrity,
and replay protection to the RTP.
V
In a VoIP network, the actual voiceconversations are transported acrossthe transmission media
using Rl Pand RTCP. or its derivatives. SRTP and cRTP. RTl' defines a standardized packet
fonnat for delivering audio and video over the Internet. RTCP isa companion protocol to RTP,
aridprovides for the delivery of control infonnation for individual RTP streams. cRTP and
SRIP were developed to enhance the use of RTP.
Datagram protocols, such as I'DP. send themedia stream as a series of small packets. This is
simple and efficient: however, packets can belost or corrupted intransit. Depending onihe
protocol and the extent of the loss, the client may be able to recover the data with error
correction techniques, may interpolate over themissing data, or may suffer a data dropout. RTP
andRTCP were specifically designed tostream media overnetworks. They are both builton
top of UDP.
2-18 Implementing Cisco Voice Communications and QoS (CVOICE! v8.0
Thissubtopic describes the RTP protocol.
Provides end-to-end deivery for real-timedata, such as voice and video
Randomly pickseven ports fromUDP port range 16384-32767
Includes the following services:
Payloadtypeidentification (codec type and mediaformat)
Allows the codec to change duringtransmission, as withfax/modem pass-
through
- Sequence numbering
* Primarilyto detect packet loss
- Measuring delay/jitter
Toplace packets inthe correct timing order (playout delaycompensation)
RTP. described in RFC3550, defines a standardized packet format for delivering audio and
video over the Internet.
RTP typically runs ontop of UDP sothat it can use the multiplexing and checksum services of
that protocol. RTP applications are typically sensitive todelays, soUDP isa better choice than
the more complex TCP. RTP does not have a standard portonwhich it communicates, The
only standard thatit obeys is that UDP communications aredone viaaneven port, and thenext
higher odd port isused for RTCP communications. Although there are nostandards assigned.
RTP commonly uses ports 16384 to 32767. The fact that RTP uses a dynamic portrange makes
it difficult for it to traverse firewalls.
The functions of RTP include the following:
Payload type identification, which identifies thetype of payload carried inthepacket, such
as codec, or media fonnat. This identifier allows the changing of codecs and data formats
while the flow is active, as is the case with fax and modem pass-through.
Sequence numbering, which monitors thesequence of arriving packets andis primarilj
used todetect packet loss. RTP does not request retransmission if a packet islost.
Time stamping, which isnecessary toplace the arriving packets inthe correct liming order.
The dejitter buffer evaluates this parameter when compensating thevariable path delay.
RTP supports both unicast and multicast transmission. Inaddition totheroles of sender and
receiver. RTP also defines the roles of translator and mixer to support the multicast
requirements.
VoIP Call Legs 2-19
Real-Time Transport Control Protocol
fhis subtopic describes the RTCPprotocol.
2-20
*al-Time Transport Control Protocol
Monitors media quality and provides control information
Provides feedback on the RTP session'
Packet count
Packet deJay
Octet count
Packet loss
Jitter (variation in delay)
Provides a separate flowfrom RTP for UDPtransport use1
Is paired with rls RTP stream
RTP stream UDP port plus 1 (odd-numbered port)
*/
IP
RTP
RTCP
*.:, --' Gatekeeper
RTCP. defined in RFC3550. is a sister protocol of RTP. RTP provides out-of-bandcontrol
information for an RTP flow. Although it is usedperiodically lo transmit control packets to
participants in a streaming multimedia session, the primary function of RTCP is to provide
feedbackon the qualitv of service (QoS) being provided by RTP.
RTCPis used for QoS reporting. It gathers statistics on a media connection, such as bytes sent,
packets sent, lost packets, jitter, feedback, and round-trip delay. Applications use this
information to adjust the transmission parameters.
There are several tv pes of RI'CPpackets: sender report packet, receiver report packet, source
description RTCP packet, goodbye RTCP packet, and application-specific RTCPpacket.
RI'CP prov ides the following feedback on current network conditions:
RTCP provides a mechanismfor hosts involved in an RTP session to exchange infonnation
about monitoring and controlling thesession. RTCPmonitors the qualityof elementssuch
as packet count, packet loss, delay, and interarrival jitter. RTCPtransmits packets as a
percentageof session bandwidth, but at a specific rate of at least every five seconds.
The R'l'Pstandard stales that the NetworkTime Protocol (NIP) time stamp is based on
sv nchronized clocks. The corresponding RTP lime stampis randomly generated and based
on data packet sampling. Both NTP and RTPare included in RTCPpackets by the sender
of the data.
RTCPprov ides a separate How from RTP for transport use by UDP. Whena voice stream is
assigned UDP port numbers. RTP is tvpicallv assigned an even-numbered port and RTCP is
assigned the next odd-numbered port. Kachvoice call has four ports assigned: RTP with RTCP
in the transmit direction and RTP with RTCP in the receive direction.
Compressed RTP
This topic describes cRTP protocol.
Compressed RTP
Maps 40-byte header to 2 (without checksum) or 4 (with checksum) bytes
most of the time
Worfcspoint-to-point, must be configured on both ends of the link
On high-speed links, processing overhead does not justify the bandwidth
savings
Algorithm:
- Establishes session context (full IP/UDP/RTP headers, few first-order
differential values, link sequence number, generation number, and a
delta encoding table)
- Session state shared between the compressor and the decompressor
After the context state is establshed, compressed packets may be sent
- Only change (delta) indicators are transmitted
cRTP on low-speed
serial links !<*76S kbiSj
RTP/RTCP Stream
The overhead introduced by packet headers is often considerably larger than the voice payload.
The overhead consists of an IP (20 octets), UDP (8 octets), and RTP header (12 octets) and
amounts to 40 bytes.
cRTP. specified in RFCs 2508, 2509. and 3545, was developed to decrease ihe sue of the IP.
UDP. and RIP headers. cRTP maps the IP/UDP/RTP header to two bytes (without checksum)
or four bvtes (with checksum).
RTP header compression is supported on point-to-poinl inlerfaces, such as serial lines using
Frame Relay. High-Level Data Link Control (HDLC), or PPP encapsulation. It is a link-local
mechanism that must be enabled on both sides of the link.
cRTP is recommended for slow-speed links of up to 768 kb/s. On faster links, the bandwidth
savings may be offset by an increase in CPU utilization on the router.
cRTP Operation
During compression of an RTP stream, a session context is defined. For each context, the
session state is established and shared between the compressor and the decompressor. The
context stale consists of the complete IP/UDP/RTP headers, a few first-order differential
values, a link sequence number, a generation number, and a delta encoding table. The context
state must be synchronized between compressor and decompressor for successful
decompression to take place.
After the context state is established, compressed packets may be sent. The compressed header
carries pointers to the respective context entities and the difference fromthe previous packet
(delta).
) 2010 Cisco Systems, Inc. VoIP Call Legs 2-21
Secure RTP
This subtopic describes SRTP protocol.
Secure RTP
- Encryption
Makes the content undecipherable for transit
Message integrity
Adds a fingerpnnt to detect tampering during transit
Message authentication
Protects the fingerprint with key to guarantee the authenticity of the
source
Replay protection
Sequencing prevents the injecting of outdated information
*
SRTPSlream
SRTP. de lined in RFC 371 I. is designed to provide encryption, message authentication and
integritv. and replay protection to the RTP data in both unicast and multicast applications.
SRTP also has a sister protocol, called Secure RTCP (SRTCP). SRTCP provides the same
securit)-related features lo RTCP as those provided by SRTP lo RTP. SRTP can be used in
conjunction with compressed RTP.
Encryption
Lnervption is the conv ersion of data into a form, called a ciphcrtcxt. which cannot be
u iderstood b\ unauthorized people. This feature is also referred lo as privacy. It ensures that
the conversation content is kept private among the endpoints. If a hacker intercepts the packets,
the hacker will not be able to decipher them. Decryption is the process of converting encrvpted
data back into its original fonn. so that it can be understood. SRTP uses the Advanced
Encryption Standard (ALS).
Authentication and Integrity
Lnervption algorithms do not secure message integritv themselves, allowing the attacker to
forge the data. SRTP prov ides the means to ensure packet integrity.
Mashed Message Authentication Code-Secure I lash Algorithm 1 (IIMAC-SIIA-I) authenticates
Ihe message and protects its integrity. Authentication provides the assurance that the VoIP
stream is coming from the authentic endpoint. and not someone impersonating the endpoint.
fhis method produces a 160-bit result, which is then truncated to 80 bits to become the
authentication tag that is then appended to the packet. The HMAC is calculated over the packet
pavload and material from the packet header, including the packet sequence number. If a
hacker tampers with the packets, the recipients will detect the tampering by verifying the
HMAC authentiealor.
Implementing Cisco Voice Communications and GoS (CVOICE] v8.0 2010 Cisco Systems, Inc.
Replay Protection
SRTP uses sequencing toprotect against replay attacks. Areplay attack is a form of
ciy ptographic attack, in which thehacker sends outdated information to force some action on
therecipient end. Toprevent suchattacks, thereceiver maintains the indices of previously
received messages, comparing them withthe index of eachnewly received message and
admitting thenew message only if it has notbeen played before. This function relies onthe
integrity protection that preventsspoofing of messageindices.
12010 Cisco Systems, Inc. VoIPCall Legs 2-23
Secure RTP Packet Format
This subtopic describes the SRTP message format.
Secure RTP Packet Format
V 1PIXI CC JM j PT | Sequence
Time Stamp
Number
Synchronization Source (SSRC) Identifier
Contributing Sources (CSRC) identifier (Options!)
RTP BdenskmfOptkwa!)
X
SRTPMKI0 Bytes for Voice
SHA-1 Authentication Tag (Truncated Fingerprint)
Encrypted Data Authenticated Data
SRTP differs from RTP onlv in the encrvpted voice payload and flic 32-bit SHA-1
authentication tag that is added to the packet. The authentication tag holds the first 32 bits of
the 160-bit SHA-1 hash digest that was computed from the RTP header and the encrypted voice
pavload ("truncated fingerprint"). The shortening of the fingerprint from 20 to 4 bytes is
considered to offer sufficient integrity protection while keeping the overhead at a minimum.
The fields used in die RTP header, such as Pav load Type, Sequence Number. Time Stamp, and
the remaining flags are carried in SRTP packets in cleartcxt. allowing the same packet
processing as with Rl P.
A>shown in the figure, the RTP packet header and the RTP payload (encrvpted voice) are
authenticated. RTP encrvption is performed before RTP authentication.
2-24 Implemerting Cisco Voice Communications and QoS (CVOICE) vB i )2010 Cisco Systems, Inc
VoIP Media Considerations
This subtopic explains some conditions for transporting VoIP overfirewalls.
VoIP Media Considerations
Firewalling
Signaling sessions use static, well-known ports
- H.323 (TCP and UDPport 1720), SIP (UDPand TCP port 5060),
MGCP (UDP port 2427), SCCP (TCP port 2000)
- Can be easily allowed through firewalls using static ACLs
* RTP/RTCP use dynamically negotiated UDP ports
- Difficultto allow through firewalls using static ACLs
- Stateful firewallsopen the RTP/RTCP ports on demand:
Works well if RTP/RTCP streams followthe signaling path
- Transmission fails when RTP and RTCP flows take different
paths from signaling
Signaling Session
VoIPconsists of twokey components: signalingand media. The signalingprotocols usestatic
portnumbers. Thedefault values are: H.323 (TCP/UDP port 1720), SIP(TCP/UDP port 5060).
MGCP (UDP/2427), SCCP(TCP/2000). Static portsallowthe firewalls to easily identify the
signaling traffic andeitherallow or blockit, depending onthe security policy.
RTPand RTCPstreams use dynamically negotiated UDPport numbers. Staticaccesscontrol
list (ACL) filters are not able to selectively allowor block certain media streams.
Stateful firewalls, such as the Cisco Adaptive Security Appliance, track the RTP port
negotiation managed by the signalingprotocol,and selectively allowihe negotiated UDPports,
if thepreceding signaling session waspermitted bythe firewall policy. All otherports remain
blocked and onlythe currently negotiated portsarc passedthrough. This technique workswell
if the RTP and RTCP sessions flow over the same firewall as the signaling messages. If the
pathsdivert, the RTP andRTCP streams will bedropped bya firewall, because that firewall
has not processed thesignaling messages andtherefore has not opened theUDP ports. To avoid
suchproblems, the network designshouldensurethat the mediastreams take the samepathas
the signaling.
VoIP Call Legs
This subtopic describes Vol Pguidelines related toprivacy.
Privacy
- IPsec protection of SRTP packets encrypts already-encrypted data
Exclude SRTP packets from IPsec protection:
To save bandwidth and computational resources
Prefer SRTP over IPsec
Less overhead
More uniform approach (covers other calls, such as from roaming users)
ESP
Header
Encryp:eS Data [Black Be.
UDP
Header
Protected Voice
Payload
In inlersite communications, the enterprise oftensecuresthe trafficexchanged between the
locations. The most common VPN technology used in such cases is IP Security (IPscc). with
Encapsulating Security Payload(ESP) as the encryption and authenticationprotocol. ESP
provides the same type of security as SRTP. Protectingthe voice media using both IPsec and
SRTP at the same time is superfluous, because it increases the overhead and consumes
computational resources without adding any significant security advantage.
If both security methods (SRTP and IPsec) are deployed in the network, SRTP is typically
recommended to secure calls for these reasons:
SRTP creates less overhead than IPsec. thus consuming less bandwidth and improving
delay,
SRTP can protect all other VoIP calls, such as from roaming users, allowing a more
uniform approach to voice security.
2-26 Implemerting Cisco Voice Communications and QoS (CVOICE) v8.Q >2010 Cisco Systems. Inc
Voice Activity Detection
This topic describes the VAD functionality.
Voice Activity Detection Overview
Builds onthenature ofhuman conversation
- One speaks, one listens
. Classifies packets into: speech, silence, and unknown
- Speech and unknown packets are sent over the network
- Packets that would carry silence arediscarded
upto 35 percent bandwidth savings
- Based on average volume of more than 24 calls
. The sound quality could be slightly degraded by VAD
- Initial after-silence sounds chopped off
Usteniig
Speaking
VAD is atechnology that builds on the nature of human conversation, where one person speaks
while others listen.
VAD classifies VoIP packets into three classes: speech, silence, and unknown. With VAD
Illed speech and unknown packets are sent over the network, silence packets are dtscarded.
VAD provides amaximum of 35 percent bandwidth savings based on an ^^*
more than 24 calls Bandwidth savings of 35 percent is asubjective figure and does not take
Z account loud background sounds, differences in languages, and other factors. The savmgs
will vary on every individual voice call or on any specific point measurement.
Note
For the purposes of network design and bandwidth engineering, VAD should not be taken
into account, especialty on links that will carry fewer than 24 voice calls simultaneousiy.
Vanous features, such as music on hold (MOH) and fax, render VAD ^^^
network is engineered for the full voice call bandwidth, all savings prov.ded by VAD are
available to data applications.
The degradation in voice quality may be noticeable when the initial sounds are:choppedôff
after arperiod of silence. In such cases, the disabling ol VAD usually solves the problem.
VoIP Call Legs
2-27
2-28
Bandwidth Savings
This subtopic provides figures .hat describe ,he bandwidth savings incurre
ngs incurred bv VAD.
bandwidth Savinqs
Codec
Codec
J speed
Sample
size
;:R*fay, tk,
e.;-wD. .
Sal
liT|#;
G711
64 kb/s
240 B
76.3 kb/s
49.6 kb/s
G 711
64 KO/s
160 B
82 4 kb/s
53.6 kb/s
LBC
13.3 kb/S
308
26.1 kb/s
17.0 kb/s
LBC
15 2 kb/s
20B
34.4 kb/s 22.4 kb/s
G729
8 kb/s
30B
20.3 kb/s 13 2 kb/s
G72S
8 kb/s
20B
26 4 kb/s
17.2 kb/s
Voll't!^","" "T i,Klr!eS I'" ta"dWidlh S0VineS ad,ievcd b>' VAR ^n lrasmii
^rîtr. v^rdihs-u,k,ne in, acco"nt ,hc en,irc mM<2 d
Implementing Cisco Voice Commu
nicatjons and QoS (CVOICE] v8.0
>2010Osco Systems. Inc
This subtopic explains the voice port settings that relate to VAD.
Music Threshold and Comfort Noise
VcxeeActivity Detection active when:
- Not disabled in trie matched VoIP dial peer
The negotiated codec supports it
Tunable voice port parameters:
Mnimal decbel level of music-on-hold
Defines loudness threshold to correctly interpret and (ransmit MOH
Local generation of comfort noise
Local telephone hears comfort noise during silence of the other end
Speaking
IPWAN
1 Ivdip (>.1 voipi-
.,ih> J htmmmmJ
< 1
VAD
Listening
VAD is enabled by default if the negotiated codec supports il. It can be disabled in the dial peer
configuration mode.
Two VAD-rclatcd parameters are configured on voice ports: comfort noise generation (CNG)
and music threshold.
Comfort noise creates subtle background noise to fill silent gaps during the conversation. If
comfort noise is not generated, the resulting silence can fool the caller into thinking the call is
disconnected instead of being merely idle. CNG provides locally generated white noise to give
the speaker the impression of background noise coming from the other end.
The music threshold specifies the minimal decibel level of music played when calls are put on
hold. The music thresholdmay be tuned to ensure that music-on-hold is correctly interpreted as
media and not classified as silence packets.
Summary
"histopic summari/es the kev points that were discussed in this lesson.
Summary
VoIP uses IP transport for call signaling and media
transmission.
Voice conversion into VoIP involves sampling, quantization,
encoding, and optional compression.
Vol P packetization affects bandwidth by defining the length of
audio put into one VoIP packet.
Voice and Video are carried in IP networks using Real-Time
Transport Protocol or its variants.
Voice Activity Detection saves up to 35 percent of bandwidth
by discarding silence packets.
2-30 Implementing Cisco Voice Communications and QoS (CVOICEI v8.0 2010 Cisco Systems. Inc
Lesson 2
Explaining H.323 Signaling
Protocol
Overview
H.323 gateways are among the most common Cisco IOS voice gateways within Cisco Unified
Communications Manager environments. H.323 gateways aretheendpoints ona LAN that
prov ide real-time, two-way communications between H.323 terminals on the LAN and other
ITU-T terminals onthenetwork. H.323 gateways canalsocommunicate withother H.323
gateways. Gateways enable H.323 terminals tocommunicate with terminals that are not H.323
terminals by converting protocols. Gateways are the point where acircuit-switched call is
encoded and repackaged into IPpackets. Because gateways function as H.323 endpoints, thev
prov ide admission control, address lookup and translation, and accounting services.
Objectives
Upon completing this lesson, you will beable todescribe the characteristics of 11.323 and
explain when touse H.323. Further, you will beable toexplain the customizable parameters
such as Fast Start, early offer, transport protocol, and interface binding. This ability includes
Describe the functional components of the H.323 environment, the functions that are
performed by a typical H.323 gateway, andthe advantages of H.323
Describe the11.323 signaling stack, the signaling messages, and H.323 call flows oncall
setup and call teardown
Describe the process of codec negotiation in H.323, the related protocols, and mechanisms
such as Fast Start
Describe how to configure an H.323 gateway
Describe howto configureH.323 interface binding, transport protocol, and howlo tune
H.323 timers
Describe majorcommands that are usedto verify an H.323 gateway
H.323 Architecture
2-32
This topic describes the 11.323 signaling protocol, its architecture, and advantages.
H.323 Architecture Ove
H.323 is a suite of protocols for voice, video, and
data with the following characteristics:
ITU standard
A mature protocol
Based on ISDN Q.931
* Vendor-neufral
* Peer-to-peer architecture
Supported on Cisco voicegateways and all Cisco Unified
Communications call control platforms
Widely deployed
H.323 is a suite of protocols that ITUdefines for multimedia conferences over LANs. It was
developed basedon ISDN Q.93 I as a protocol to provide IPnetworks withtraditional
telephony functionalilv. H.323 is a mature, vendor-neutral protocol that is currently themost
widely deployed standards-based voice andvideoconferencing standard for packet-switched
networks.
H.323 isa peer-to-peer protocol inwhich each gateway plays anequal part inthe signaling
process and must maintain its owndial planto makecall forwarding decisions. This
characteristic differentiates 11.323 from server-client signaling protocols such as Media
Gateway Control Protocol (MGCP), where thegateway registers on ihecall agent toreceive
further instructions. 11.323 is supported on all Ciscovoicegateways and all Cisco Unified
Communications call control platforms.
H.323 describes an infrastructure oflerminals, common control components, services, and
protocols that are used for multimedia (voice, video, and data) communications.
An H.323 gatewav is anoptional type of endpoint that provides interoperability between H.323
endpoints and endpoints that are located on a Switched CircuitNetwork(SCN). suchas the
public switched telephone network (PSTN) or anenterprise voice network. Ideally, the gateway
is transparent to both the H.323 endpoint andthe SCN-based endpoint.
*.
H.323 Advantages
This subtopic explains the advantages of H.323.
H.323 Advantages
Self-sufficient dial plan per gateway
- Call routing configuration canbemore specific than onCisco
Unified Communications Manager
- Noneed forextra call-routing configuration when using Cisco
Unified SRST
Translations defined per gateway:
- Regional conditions can be metwithin multisite deployments
Nodependencyon the CiscoUnified Communications Manager
- Supportfor morevoiceinterface types than with MGCP
Support for ISDNNFAS
Enhanced fax support and call preservation
There are several advantages to usingH.323 gateways as voicegateways:
Self-sufficient dial plan per gateway: It enables processing thecall routing locally
without relying on a call agent, as is thecasewithMGCP.
Call-routing configurationcan be more specific than on CiscoUnified
Communications Manager: Cisco IOS gateways enable translating and matching tothe
called number andthecalling number, which canimprove call routing. CiscoUnified
Communications Manager matches only thecalled number. Forexample, this difference
enablescall routingfrom unwanted peopleto a special destination.
There is no needfor extra call routingconfigurations that are related to CiscoUnified
Sun ivableRemote Site Telephony(SRST): Because thecall routing configuration is
donedirectly onthegateway, noadditional dial planis required for SRST.
Translations can bedefined per gateway: Thissupports regional requirements such as
calling party transformations or special number formats. All incoming and outgoing calls
can be translaled directly on the gateway to meet the internally usednumber format.
There is nodependency on the Cisco Unified Communications Manager: Because the
configuration is perfomied on thegateway andthe H.323 umbrella is a pecr-to-peer
protocol, there is no dependence onsoftware versions and feature sets ofother signaling
components.
Morevoice interface types are supported: Because theCisco Unified Communications
Manager docs notneed tocontrol the interface cards within H.323 environments, more
interface cards are supported whenyou use H.323 rather than MGCP.
ISDN Non-Facility Associated Signaling (NFAS) is supported: The H.323 gateway
signalingprotocol supports NFAS, which MGCP does not.
VoIP Call Legs 2-33
Enhanced fax support: Fax support is better on 11.323 gateways than on MGCP gatewav s
because 11.323 supports T.37 and T.38. An H.323 gateway can route afax direct inward "
dialing (DID) number directly to a Foreign Fxchange Station (FXS) port on the gatewav.
Knhanced call preservation: Call preservation is useful when agateway and its
communicating peer(typically a Cisco Unified IPphone) arecollocated while thecall was
signaled o\er a Cisco Unified Communications Manager resident inanother site. When the
WAN connectiv it; fails, the media connection between the gateway and the phone will
remain active because of the call preservation enhancements.
2-34 Implementing CiscoVoice Communications and QoS (CVOICE! v8.0 2010CiscoSystems. Inc
3te
1w
This subtopic describes the network components ofan H.323 environment.
The figure shows some typical terminal devices inan H.323 network.
An H.323 network includes the following components:
Terminals: H.320(ISDN). H.323, H.324(plainold telephone service[POTS])
Gateways
Gatekeepers
Multipoint control units
Cisco Unified Border Elements
H.323 Terminals
An H.323 terminal is anendpoint thatprovides real-time voice (andoptionally, video anddata)
coninuinications with another endpoint, such as an H.323 terminal or multipoint control unit.
The communications consist of control, indications, audio, moving color video pictures, or data
between the two terminals. A terminal may provide the following:
Audio only
Audio and data
Audio and video
Audio, data, and video
The terminal canbe a computer-based videoconferencing systemor other device.
An H.323 terminal mustbecapable of transmitting andreceiving voice that is encoded with
G.711 (a-law andmu-law). andmay support otherencoded voice formats, such as G.729 and
G.723.1.
VoIP Call Legs
H.323 Gateways
This subtopic describes H.323 gateways.
H.323 Gateways
H.323 gateways perform these services:
Translation between audio, video, and data formats
Conversionbetween call setup signals and procedures
Conversion between communication controlsignals and
procedures
H 323 Device
The figure shows a gateway connectingan H.323 device, and a terminal that is not an 11.323
terminal, such as an analog telephone. The H.323 device can be a 11.323 terminal, multipoint
control unit, gatekeeper, or another H.323 gateway.
(jatevvays allow H.323 dev ices locommunicate with devices that arc running other protocols,
"fhe; provide protocol conversion between thedevices that arerunning different types of
protocols. Ideally, the gateway is transparent to boththe H.323 endpoint and the non-H.323
endpoint.
An 11.323 gateway performs these services:
Translation between audio, video. and data formats
Conversion between call setup signals and procedures
Conversion between communication control signals andprocedures
>20!0Cisco Systems. Inc
H.323 Gatekeepers
This subtopic describes the H.323 gatekeepers.
H.323 Gatekeepers
An 11.323 gatekeeper provides address translation and access control for H.323 terminals,
gateways, and multipoint control units. Gatekeepers are optional nodes that manage endpoints
inan H.323 network. The endpoints communicate with thegatekeeper using theRegistration.
Admission, and Status (RAS) protocol.
Endpoints attempt to register with agatekeeper on startup. When they wish to communicate
with another endpoint. they request admission toinitiate acall. If the gatekeeper decides that
the call can proceed, itreturns adestination IP address tothe originating endpoint. This IP
address may not bethe actual address of the destination endpoint, but an intermediate address,
such asthe address of a proxy or a gatekeeper thatroutes call signaling.
When a gatekeeper is included, it performs these functions:
Address translation: Converts an alias address to an IP address
Admissioncontrol: Limits accessto networkresourcesbasedon call bandwidth
restrictions
Bandwidthcontrol: Responds tobandwidth requests andmodifications
Zonemanagement: Provides services to registered endpoints
The gatekeeper may also perform these functions:
Call authorization: Rejectscalls basedon authorization failure
Bandwidth management: Limits the number ofconcurrent accesses toIPinternetwork
resources (Call Admission Control |CAC])
Call management: Maintains a record of ongoing calls
H.323 gatekeepers arecovered in more detail ina latermodule.
VoIP Call Legs 2-37
H.323 Multipoint Control Units
This subtopic describes the multipoint control units.
H.323 Multipoint Control Uni
H 323 -t 323
Termina' Termina
i H323
"^ Terminal
Amultipoint control unit isan endpoint onthe network that allows three or more endpoints to
participate in a multipoint conference. It controls and mixes video, audio, and data from
endpoints tocreate a robusl multimedia conference. Amultipoint control unit may also connect
two endpoints in a point-to-point conference, which may later develop into a multipoint
conference.
2-38 implementingCisco VoiceCommunications and QoS (CVOICE) v8.0
This subtopic explains the different types ofmultipoint conferences.
666666
s
-3r
V
r
vitteo
Audio
Multipoint conferences rely on asingle multipoint control unit to coordinate the membership of
a conference. Each endpoint has an H.245 control channel connection tothe multipoint control
unit. Fither the multipoint control unit orthe endpoint initiates the control channel setup. H.323
defines three main types ofmultipoint conferences: centralized, distributed, and adhoc:
Centralized multipoint conference: The endpoints must have their audio, video, ordata
channels thatareconnected toa multipoint processor (MP). TheMP performs mixing and
switching ofthe audio, video, and data, and if the MP supports the capability, each
endpoint canoperate ina different mode.
Distributed multipoint conference: The endpoints do not have a connection toanMP.
Instead, endpoints multicast their audio, video, and data streams to all participants in the
conference. Because an MPis not available for switching andmixing, anymixing of the
conference streamsis a function of the endpoint, andall endpoints must use the same
communication parameters.
Adhoc multipoint conference: An ad hoc multipoint conference isa hybrid situation, in
which theaudio andvideo streams are managed by a single multipoint control unit but
where onestream relies on multicast (according tothedistributed model) andtheotheruses
the MP (as inthe centralized model). Any two endpoints inacall can convert their
relationship into a point-to-point conference. When the point-to-point conference is created,
other endpoints become part ofthe conference by accepting an invitation from a current
participant, orthe endpoint can request tojoin the conference.
VoIP Call Legs 2-39
H.323 Regional Requirements Example
Ihis subtopic prov ides an example of regional requirements.
Translate
and route
Seasonal
caNtng number I us / Maria\
Ir thisscenario. Maria with thenumber 91721611 I from Spain calls Alice in the United States
and Frank in Gennanv. The procedure enabling Alice and Frank tocall back Maria using their
missed call list is as following:
1. Maria places acall loAlice in the United Stales. Because the International Direct Dialing
(IDD) prefix for Spain is34. the number that issent out asthe calling party by the Spanish
provider is 3491721611! with inlenialional" as thetvpe of number (TON),
2. When the call arrives onIhe U.S. gatewav. the calling party number (349172161 11) is
translated tomeet the common dialing regulations ofthe United States: 011 is prepended as
the international dialing prefix, and a leading 9 isprepended as the access code for external
calls from the company network.
3. Ihemissed calls liston Alice's phone displays a call from 901134917216111, andshe will
be able to reach Maria hv using the callback feature.
4. Maria places a call to Frank inGermany. The calling parly number for Maria is
34917216111 with the international TON.
5. When the call arrives on[he German gatewav. ihecalling party number (349172161 I I) is
translated tomeet the dialing regulations ofGermany: 00 is prepended as the IDD prefix,
and a leading 0isprepended as the access code for external calls from the company
network.
6. The missed calls list on Frank's phonedisplaysa call from 000349172161 I Land he will
be able to reach Maria bv using the callback feature.
2-40 Implementing CiscoVoice Communications and QoS (CVOICE) v8.0
H.323 Call Flows
This topic describes the H.323 signaling stack and explains how H.323 establishes and
terminates calls.
The figure shows theelements of anH.323 terminal andhiglilights theprotocol infrastmcture
of an H.323 endpoint. H.323 is considered an"umbrella protocol" because it defines all aspects
of call transmission, fromcall establishment to capabilities exchange to network resource
availability. H.323 defines these protocols:
11.225 for call setup: fhe call-signaling function allows anendpoint to create connections
withotherendpoints. Thecall-signaling function defines call setupprocedures that are
basedon the ISDN ITUQ.931 protocol, whichallowsinteroperability withthe PSTN and
Signaling System 7 (SS7).
H.225 for Registration, Admission, and Status (RAS) control:The RAS signaling
function uses a separate signaling channel toperform registration, admissions, bandwidth
changes, status, anddisengage procedures between endpoints anda gatekeeper.
H.245 for capabilities exchange: The H.245 control channel is separate from thecall-
signalingchannel and is responsible for these functions:
Logical channel signaling: Opens andcloses the Real-Time Transport Protocol
(RTP) or Real-TimeTransport Control Protocol (RTCP) media streams.
Capabilities exchange: Negotiates audio, video, andcodec capabilities.
Master or responderdetermination: Determines which endpoint is a master and
whichis a responder. It is usedto resolveconflictsduringthe call.
Mode request: Requests a changeinmode, or capability, of the mediastream.
VoIP Call Legs 2-41
H.323 Slow Start Call Setup
fhis subtopic illustrates the slow start call setup between two 11.323 gateways.
H.323 Slow Start Call Setui
PSTN/
Pnvat-
Voica
1. Initiate Call I
PSTN/ S
Private
H225.'Q931
Call Setup
6. Ringback Tone |
H245
Capabilities
Negoliation
12. Media (RTP) {
The diagram shows an H.323 slow startcall setupexchange between twogateways. Thesame
procedure is used when one or both endpoints arc H.323 terminals:
stepi
Step 2
Step 3
Step 4
Step 5
Step 6
Step 7
Step8
Step 9
Step 10
Step 11
Step 12
An endpoint initiates a call.
Originating gateway initiates an H.225 session with thetenninating gateway on TCP
port 1720. Originating gateway determines terminating gateway address from its
local configuration.
Tenninating gateway acknowledges theCall Setup with theCall Proceeding
message.
Terminatinggatewav sends the ringing signal lo the recipient telephone.
Terminating gateway notifies theoriginating gateway about Iheringingwiththe
Alerting message.
Originating gatewav signals theringback tone totheoriginating endpoint.
Recipient takes the phone off-hook,
Tenninatinggatewav sendsthe Connectmessage to originating gateway.
The endpoints open another channel for the H.245 control function, 'fhe If.245
control function negotiates capabilities.
Ihe H.245 control function determines the master/slave roles to resolve potential
conflicts.
The H.245 control function exchanges openlogical channel messages thai describe
RTP flows.
Ihe gatewav s start transmitting mediaover the RTP channels and exchanging call
qualitv statistics using RTCP.
>2010 Cisco Systems, Inc
H.323 Slow Start Call Teardown
This subtopic illustrates theprocess of H.323 slow start call teardown.
H.323 Slow Start Call Teardown
H24S
Teardown
Negotiation
H 225 Call
Teardown
%.Close Logical Channel
3. Close Logical Channel ACK
4. End Session Command
5. End Session CommandACK
G. Release Complete
1. Hangup
The diagram shows a slowstart H.323 call termination between two gateways:
Step 1 One communicating party hangs up. This example shows the endpoint behind the
terminating gateway, but this procedure would be mirrored if the endpoint behind
the originating gateway hung up.
Step 2 Terminating gateway sends Close Logical Channel message to the originating
gateway.
Step 3 Originating gateway acknowledges the message.
Step 4 Tenninating gateway sends End Session Command message to the originating
gateway.
Step 5 Originating gateway acknowledges the message.
Step 6 Terminating gateway sends the Release Complete message to the originating
gateway.
12010 Cisco Systems, Inc VoIP Call Legs 2-43
H.225 RAS Call Setup
This topic describes the call setup procedure using RAS ina gatekeeper-controlled
cm ironmenl.
H.225 RAS Call Setup
H 225/Q931
Call Setup
H245
Capabilities
Negotiation
PSTN,'
Pnvate
H225
RAS
H225
RAS
10. Rtng&acK Tone
16. Media (RTP)
H 323 Gatekeeper
ARQ - Admission Requesl
ACF -Admission Confirm
Ihe diagram shows an H.323 basic call setup exchange between two gateways. The same
procedure is used when one or both endpoints are 11.323 terminals:
Step 1 An endpoint initiates a call.
Step 2 Originating galewav initiates an H.225 sessionwiththegatekeeper on registered
RAS port TCP/I 719. The gatekeeper listens on TCP port 1718 for discovery
messages, and the discovery process must be completed before the gateway can send
RAS messages to the gatekeeper. The gateway sends the Admission Requesl (ARQ).
Step 3 1he gatekeeper returns Admission Confirm (ACT) that includes the IP address of the
tenninating gatewav.
Step4 Originating gatewav initiates an 11.225 sessionwiththe terminating gateway on port
TCP/1720 using the H.225/Q.931 Call Setup message.
Steps Tenninatinggalewav sendsARQlo the gatekeeper (TCP/1719) requesting
permission to accept the call.
Step 6 Gatekeeper returns ACT7 to the terminating gateway, granting permissionto accept
the call.
Step 7 terminating gatewav acknowledges the Call Setup with the Call Proceeding
message to originating gatewav.
Step 8 Tenninating gateway sends the ringing signal lo the recipient telephone.
Step 9 Terminating gatewav notifies the originating gateway about the ringing with Ihe
Alerting message.
Step 10 Originating gatewav signals the ringback lone to the originating endpoint.
Step 11 Recipient takes the phone off-hook.
Implementing Cisco Voice Communications and QoS (CVOICE) v8 0 )2010 Cisco Systems. Inc
Step 12 Terminating gateway sends the Connect message to originating gateway.
Step 13 The endpoints open another channel for the H.245 control function. The H.245
control function first negotiates capabilities.
Step 14 The H.245 control function determines the master/slave roles toresolve potential
conflicts.
Step 15 The H.245 control function exchanges open logical channel messages that describe
RTP flows.
Step 16 The gateways start transmitting media over the RTP channels and exchanging call
quality statistics using RTCP.
12010 Cisco Systems. Inc. VoIP Call Legs 2-45
H.225 RAS Call Teardown
This subtopic describes the call teardown procedure inan H.323 network with gatekeeper.
H 245
Teardown
Negotiation
H 225 Call
Teardown
PSTN/
Piwat*
Voic
H 323 Gatekeeper
H323 ^gr H323
Gateway ^^0 Gateway
Jf>.__.. ._._ i>^^
pstn; /
Private .-
1. Hangup
2. Close Logical Channel
3. Close Logical Channel ACK
4. End Session Command
*
5. End Session Command ACK
6. Release Complete
7a. DRQ
8a. DCF
7b. DRQ
8b. DCF"
DRQ - Disengage Request
DCF - Disengage Confirm
The diagram showsan H.323 call lenninalion between two gateways that are registered to a
galekeeper:
Step 1 A communicating party hangs up.
Step 2 Terminating gatewav sends Close Logical Channel message to the originating
gatewav1.
Step 3 Originatinggatewav acknowledges the message.
Step 4 1erminating gatewav sends find Session Command message to the originating
gatew av,
Step 5 Originating galewav acknowledges the message.
Step 6 Tenninating gatewav sends the ReleaseComplete message to the originating
gateway.
Step 7 Bothgateways send Disengage Request (DRQ) messages to the gatekeeper.
Step 8 The gatekeeper replies to both DRQs with DisengageConfirm(DCF) messages.
Implementing Cisco Voice Communications and QoS (CVOICE) i/8.0 2010 Cisco Systems, Inc
Codecs in H.323
This topic describes the process ofcodec negotiation in an H.323 environment.
Negotiation Messages
TCS message negotiates media channel parameters
* OLC message opens a logical channel for mediatransport
Possible replies: Acknowledge, Reject, Confirm
Exchanged bydefault in the H.245signaling phase
Master/slave determination settles conflicts
- If slave tnes to open incompatible flow, master rejects it
H.323 Fast Start feature embeds this information in H.225
call setup and subsequent messages
Terminal Capability S#t
Codec(s) VAD
The H.245 call control performs three functions when a call is beingset up:
Capability negotiation isthe most important 11.245 function. Itenables devices to
communicate without having priorknowledge of thecapabilities of theremote entity. It
negotiates audio/video/text codecs, additional parameters such as voice activity detection
(VAD). andenables real-time dataconferencing. The capabilities arcoffered using
Terminal Capabilities Set (TCS) messages, and answered using anAcknowledge, Reject, or
Confirm.
Master/Slave determination occurs after the first TCS message is sent. H.323 attempts to
dctenninewhichdeviceis the "master" and whichis the "slave." The masterof a call settles
all "disputes" between thetwodevices. Forexample, if the slave attempts toopenan
incompatible media flow, the master takes the action toreject the incompatible flow. The
determination principle selects theendpoint with thelarger terminal type value as master.
There arefour terminal types (ordered from thehighest tothe lowest value): multipoint
control unit, gatekeeper, gateway, and terminal. If theterminal type values arethe same,
the master isset totheendpoint with the larger statusDeterminationNumber, which isa
random number that is generated byeachparty, in therange from 0 to 224-1.
Logical Channel Signaling occurs after capabilities are exchanged and master/slave
determination iscompleted. The devices open media flows, referred toas"logical
channels." This isdone bysending anopen logical channel (OLC) message thatcarries the
RTP/RTCP ports and receiving an acknowledgment message. Upon receipt of the
acknowledgment message, anendpoint may then transmit audio or video totheremote
endpoint.
VoIP Call Legs 2-47
Negotiation in Slow Start Call Setup
This subtopic details the negotiation during aslow start call setup without agatekeeper.
!Qotiation in Slow
H 225,'Q 931
Call Setup
H245
Capabilities
Negotiator!
PSTN/
Pnvatft
Voice
1.
3. CallProceeding
5. Alerting
8. Connect
* *~******~
9. TCS Request
JO. Waster/Slave Request
11 TCS Request
12. Master/Slave Request
13. TCS + Master/Slave ACK
15, Master/Slave ACK
16. OLC Requesl
17. OLC Request
18. OLC ACK
19. OLC Response
20. Media (RTP)
4 Ring Called Party
7. Answer Call
TCS- Terminal Capabiir
(Codec. VAD)
OLC -Open LogicalClu
(RTP/RTCP pour
This diagramprovides a detailed description of all H.225and H.245 messages that are
exchanged during call setup without a gatekeeper. It shows thattheH.245 exchange is triggered
by the tenninating gatewav in Step 9. The capability negotiation and master/slave
determination is performed in the first six H.245 messages (Steps9 through 15). After the
capabilities have been continued andthemaster determined, theoriginating gateway starts the
logical channel signaling phasethat consists oflbur messages (Steps 16through 19). When the
OLC messages (with RTP/RTCP port numbers) have been confirmed, thegateways start
streaming voice media.
Implementing Cisco Voice Communications and QoS (CVOICE! v8.0 2010 Cisco Systems, Inc
H.323 Fast Connect
This subtopic describes the H.323 Fast Connect (Fast Start) feature.
H.323 Fast Connect
H225 Call Setup message carries multipleH.245TCSrOLC
combinations, based on the number of codecs
| 3. Call Proceeding
H 225 Call Proceeding message carries confirmation for one
TCS variant and OLC information
6. Ringback Tone
5. Alerting
_8. Connect
4. Ring CalledParty
j 7, Answer Call
9. Media (UDP)
The figure shows anH.323 setup exchange thatuses theFast Connect abbreviated procedure
available in H.323 version 2. The Fast Connect (Fast Start) procedure reduces the number of
round-trip exchanges and achieves the capability exchange and logical channel assignments in
one roundtrip. Fast Connectis widelysupported inthe industry.
The Fast Connect feature occurs in these steps:
Step 1 An endpoint initiates a call.
Step 2 The originating gateway initiates an 11.225 sessionwiththe destination gateway on
registered TCPport 1720. The Call Setupmessage is combined withthe H.245
control channel and includes a set of capabilities and logical channel descriptions.
The numberof these proposalsdepends on the numberof codecsthat are supported
by the originating gateway.
Step3 Terminating gateway responds usingthe Call Proceeding message that carries the
confirmation for one TCS variant and includes the OLC information about the
RTP/RTCP port numbers.
Step4 The remaining H.225 exchangefollows the same patternas in the standardcall setup
procedure, afterwhich the RTPmedia andRTCP monitoring channels start.
12010 Cisco Systems, Inc VoIP Call Legs 2-49
H.323 Early Media
This subtopic describes the H.323 F.arly Mediafeature.
H 225 Cafl Setup message carries multiple H 245 TCS/OLC
combinations based on Ihe number of codecs, and requests
Early Media
3. Call Proceeding
H 225 Call Proceeding message carries confirmation for one
TCS/OLC vanant and confirms Early Media
PSTN; S
Private;;
7. Ringback Tone
' ?;..! l"9
4. Early Media allows streaming
of media (announcements, MOH)
betore the call is accepted
5. Ring Called Parly
8. Answer Call
Ihe Early Media feature buildson the Fast Connect exchange. Bothgateways negotiate the
capabilities, suchas codecs, and the RTP/RTCP port numbers withinthe first two messages
(Call Setupand Call Proceeding). Whenthe Farly Mediais also negotiated, they openthe
media channels before any other H.225 messages are exchanged.
F.arly Media allows sending of media fromthe called parly or an application server to the
caller, prior to the call being accepted. F.arly Media is usually sent fromthe PSTNand carries
ringing tones or announcements. If no audio infonnation is available for transmission before
the call is accepted, the media streams carry silence.
An example of F.arly Media is the streaming of announcements that cell phone operators allow
their subscribers to customize. When a cell phone owner records their own announcement, it is
playedwhenever the extension is calledand the cell phoneis ringing. If that call travelsover an
IP network using 11.323 signaling. H.323 Farly Media is used.
This topic describes how loconfigure H.323 gateway functionality onCisco IOS routers.
Configure VoIP dial peers
- Dial peers default to H.323 protocol
H.323 enabled by default
router(conflg}#
dial-peer voice tag voip
Creates a VoIP dial peer
Enters configuration mode for parameters such as session target,
destination-pattern, incoming called-number, answer-address
ACisco voice gateway must have at least oneVoIP dial peertoactas an H.323 originating
gateway. The default protocol of a VoIP dial peerisset to H.323. Therefore, thegateway will
use H.323 tosignal any calls thatarematched bythe outbound VoIP dial peer with the default
protocol.
ACisco gateway is, bydefault, enabled toact as an H.323 terminating gateway. When an
H.323 call is receivedon that gateway, evenwhen no dial peers exist, the gatewaytries to use
thedefault dial peertomatch theincoming setuprequest. If VoIP dial peersexist,thegateway
tries to find the inbounddial peer using the commands incoming called-number, answer-
address, and destination-pattern (in this order).
H.323 serviceis an integral part of the VoIPserviceand cannot be controlled separately from
the VoIPservice. VoIPservice is enabledbydefault and can be disabledby the administrator.
To disable or re-enable the VoIP service, you must enter the VoIP service configuration mode
using thevoice servicevoip global configuration command. TheVoIP services areenabled by
default and can be disabled usingthe shutdown command. The forced optioncauses the
gateway toimmediately terminate all in-progress calls. Disabling the VoIP service affects all
VoIP signaling protocols and media transmissions.
The dial-peer voicecommand is used to define dial peers, including VoIP dial peers. An H.323
gatewav needs VoIPdial peers to makeVoIPcalls usingH.323.
VoIP Call Legs 2-51
H.323 Gateway Configuration Example
This subtopic presents an H.323 gateway configuration example.
H.323 Gateway Configuration Exai
R1
1/0/0
1001
dial peer voice 1 vo
IP
inc rifling ca lled-nuniber
dial peer v ;ce 10 p =>tS
das initio -pattern 1001
pore 1/0/0
dial peer v ice 20 v nip
destinatio -pattern 200.
Bes
ion target ipv4 10.2.1.1
IP WAN 2002
dial-peer voice 1 voip
incoming called-number .
port 1/0/0
port 1/0/1
session target ipv4:10.1.1.1
Ihe figure shows two 11.323 gateways that are configured with the dial peers that allow H.323-
based calls between two network locations. The VoIP dial peers use 11.323 by default. H.323
signaling messages are transported by default over TCP. They use the destination IP address
that is specified in the dial-peer session target command. The source address is taken from the
outgoing interface toward that session target (the routing table points over the outgoing
interface to the destination address).
VoIP service is enabled by default, and therefore does not appear in the configuration. It could
be disabled using the shutdown command in voice service VoIP configuration mode.
Customizing H.323 Gateways
This topic describes howto customize H.323gateways.
H.323 Gateway Tuning Overview
Tuning Options:
H.323 session transport
Source IP address
H.323 timers
-- H.225 settings
The most common H.323 customization tasks include the following:
Defining the session transport protocol: TCP or UDP
Selecting a sourceIP address by bindingihe gateway functionality to a networkinterface.
Tuning H.225 timers
) 2010 Cisco Systems. Inc. VoIP Call Legs 2-53
H.323 Session Transport
This subtopic explains how to set the H.323 transport protocol.
H.323 Session Transpos
router(config]#
voice service voip
Enters the VoIP service configuration mode
router(conf-vol -serv)#
Enters H.323 configuration mode
Accessed from voice service VoIP configuration mode
No default behavior or values
no h323 command removes all commands in H.323 mode but does
not disable H.323 service
router(conf-serv-h3231S
session transport udp
Defines the H.323 session transport method
Configured in H 323 mode
Default: TCP
To customize the H.323 gateway parameters. \ou will enter ihe VoIP service configuration
mode using the voice service voip global configuration command
From the voice service VoIP configuration mode, you can enter H.323 configuration mode
using die h323 command. TTie h323 command does nol have a default behavior or values. The
no h323 command does not disable the H.323 service but only removes all commands that
were previously configured in the H.323 configuration mode.
You can change the H.323 transport protocol using the session transport udp command in the
H.323 configuration mode. To change the transport back to the default TCP setting, issue the
n session transport udp command. UDP session transport allows the shortest call setup time,
theoretical!) in as few as 1.5 round trips. TCP lakes longer due lo its overhead and
acknowledgment exchange, but guarantees packet delivery. UDP may be chosen if
communicating with third-partv devices with UDP support.
2-54 Implementing Cisco Voice Communications and QoS (CVOICE! v8.0 2010 Cisco Systems. Inc.
This subtopic explains how totune the idle connection timer and configure the interface
binding feature.
router(conf-aerv-h323)#
h.225 timeout tcp call-idle {value | never}
Sets the idle call connection timer
By default 10 seconds
router(config-if)#
h323-gateway voip bind srcaddr ip-address
Sets the source IP address for outgoing H.323 traffic
Affects H.225, H.245, and RAS messages
J
To tune the H.225 idle call connection timer, use the h225 timeout tcp call-idle command in
the H.323 configuration mode. The default idle call connection timer is 10seconds.
To configure the interface binding feature, issuethe h323-gateway voip bind sreaddr
command indie interface configuration mode. It musl be the interface withwhich the H.323
gateway service should beassociated. The command points toanIPv4 or IPv6 address of that
interface. The address will be used as the source IP address for all outgoing H.323 traffic,
including H.225. H.245. and RAS signaling.
12010 Cisco Systems, Inc VoIP Call Legs
H.225 Timers
This subtopic explains how to tune 11.225 timers.
router(coofig)#
voice class h32 3 h323_cla3B_tag
Enters Ihe H.323 class configuration mode
router(config-class)#
h225 timeout tcp establish
* H 225 TCP timeout value (default 15 sec.
router(config-class)#
h22 5 timeout setup
Response timeout value for outgoing setup messages (default 15 sec)
router(con fIg-dial-peer) #
voice-class h32 3 h323_voice_claaa tag
Attaches a voice class to a dial peer
lo tune H.225 timers, create a H.323 voice class using the voice class h323 command. Ihe
voice class is identified using a tag. In the H.323 voice class configuration mode, you can tune
these timers:
The h225 timeout tcp establish command defines the timeout, alter which the H.225 TCP
session times out if the gatewav does not receive a response. Thistimeout should be
shortened if a backup tenninatinggatewav exists, so that Iheoriginating gateway does not
have to wait the default 15seconds before contacting the backup device. A timeout of 3
seconds is recommended if the gateway communicates with a Cisco Unified
Communications Manager cluster with multiple redundant servers.
The h225 timeout setup defines the response timeout value for outgoing Call Setup
messages. Its default value of 15 seconds works well in most cases.
Finally, the H.323 voice class must be associated with dial peers. This association is configured
with the voice-class h323 command.
2-56 Implementing Cisco Voice Communications and OoS (CVOICE) v8 0 >2Q10 Cisco Systems, Inc
*r*
H.323 Gateway Tuning Example
This subtopic presents anexample of H.323 gateway tuning.
H.323 Gateway Tuning Exampl
Tuning Example
LoopbackO: 10.1 1 1
interface LoopbackO
ip address 10.1.1.1 255.25S.255.2SS
h323-gotei'ay voip bind ercaddr 10.1.1 1
voice service voip
h323
session transport tcp
voice class h323 10
225 timeout tcp establish 3
voice-class h323 10
destination-pattarn 200.
session target ipvA: 10.2.1.1
voice-class b323 10
session target ipv4: 10.3.1.1
preference 1
Physical interface redundancy
- Interface binding
Fast fallback to backup peer
- Timer set to 3 seconds
The figure shows the configurationof these features:
Interfacebinding: Thegateway usesthe 10.1.1.1 address for all outgoing H.323 packets.
The gateway uses tworedundant WAN interfaces andthe interface binding decouples
H.323 signaling from the physical path.
Transport protocol: Transport protocol is set to TCP. This command will not showinthe
configuration, as it is thedefault setting.
H.225 TCP establish timeout: The TCP establish timeout is shortened to 3 seconds to
speed up fallback tothebackup gateway if theprimary fails. Theprimary gateway
(10.2.1.1) is reached over the dial peer 1withthe best preference 0 {notshown because it is
the default value). The dial peer 2 with preference 1 points to the secondary gateway
10.3.1.1.
VoIP Call Legs 2-57
Verifying H.323 Gateways
This topicdescribes how to verify 11.323 gateways.
Verifying H.323 Gateways
Router# show gat eway
H.323 ITU-T Version: 4.0 H323 Stack Version: 0.1
B. 323 service is up
Thi9 gateway is not registered to any gatekeeper
Alias lis; (CL1 configured! is empty
Alias list {last RCF) is empty
H323 isssurce thresholding is Disabled
Use the show gateway command lo verify that the H.323gateway is operational and lo displav
the current status of the gateway.
The sample output shows the report that appears when the gateway is not registered with a
gatekeeper.
2-58 Implementing Cisco Voice Communications and QoS (CVOICE) v8 0 >2010Cisco Systems, Inc.
Summary
This topic summarizes the key points that were discussed inthis lesson.
Summary
H.323is a widely supported peer-to-peer VoIP signaling
protocol.
H.323signaling occurs intwostages: H.225 and H.245.
H.323 fast-start shortens the call setup exchange.
* H.323gatewayuses dial peers to reach other devices.
H.323gateway can be configuredto use UDP transport,
a specific IP address, and modified timers.
H.323operationscan be verified usingshow and debug
commands.
VoIP Call Legs
Lesson 3
Explaining SIP Signaling
Protocol
Overview
SessionInitiation Protocol (SIP) is one of the most important voicesignalingprotocols within
serv ice provider VoIP networks and issupported bymost IPtelephony system vendors. As
such, it is an ideal protocol for interconnecting different VoIP systems andnetworks. An
understanding ofthefeatures and functions of SIP components, and the relationships that the
components establish with each other, is important inimplementing a scalable, resilient, and
secure SIP environment. This lesson describes howto configure SIP and explores the features
and functionsofthe SIP environment, including its components, how these components
interact, and how to accommodate scalability and survivability.
Objectives
Upon completing this lesson, you will beable todescribe the characteristics ofSIP. and explain
when to useSIP. Further, youwillbe ableto explain thecustomizable parameters suchas early
and delayed otTer. transport protocol, and interface binding. This ability includes being able to
Describe SIP and its related standards, the functional and physical components of a SIP
network, and the advantages of SIP
Describe SIPsignaling messages andthethree models of SIPcall setup: direct, using a
proxy server, and using a redirect server
Explain SIPaddress formats, address registration, andaddress resolution
Describe theprocess of codec negotiation in SIP, therelated protocols, andthe Early Offer
mechanism
Describe the commands that are used to configure basic SIP functionality on Cisco IOS
gateways
Explain themajorSIPfeatures thatsupport ISDN anddescribe howtoconfigure them
Explain how to configure SIP support for SRTP
Describe howto configure SIPinterface binding,transport protocol, and SIP timers
Describe key commands that are usedto monitorand verifySIPgatewayoperation
SIP Architecture
"I his topic describes the Session Initiation Protocol, ils features, related standards, and
architecture.
SIP Overview
* Still evolving IETF standard
Creales, modifies, and terminates multimedia sessions with one or
more participants
* Leverages various standards: RTP, RTCP, HTTP, SDP, DNS, SAP,
MGCP, and RTSP
* Supported on Cisco voice gateways and Cisco Unified IP phones
that have SIP firmware
Peer-to-peer architecture
User agent client (UAC) initiates SIP requests
User agenl server (UAS) returns SIP responses
Phones, gateways, and Cisco call control devices can be UAs
- UsesASCII text-based messages
The Internet Engineering Task Force fIETF) developed SIP as an alternative to FI.323. SIP is a
common standard that is based on the logic ofthe World Wide Weband very simple to
implement. It is widelv used with gateways and proxy servers within service provider networks
for internal and end-customer signaling. Eikc other VoIP protocols, SIP is designed to address
the functions of signaling and session management within a packet telephony network.
SIP operates on the principle of session invitations lhal arc based on an HTTP-like request and
response transaction model. Each transaction consists of a request that invokes a particular
method, or function, on the server and at least one response. Through invitations, SIP initiates
sessions or invites participants into established sessions. Descriptions of these sessions arc
advertised by am one of several means, including the Session Announcement Protocol (SAP)
defined in RFC 2974. The SAP incorporates a session description according to the Session
Description Protocol (SDP) defined in RFC 2327.
SIP uses other IETF protocols to define other aspects of VoIP and multimedia sessions: for
example. URLs for addressing. Domain Name System (DNS) for service location, and
Telephony Routing over IP (TRIP) for call routing.
2-62 Implementing Cisco Voice Communications and QoS (CVOICE) v8 0 ) 2010 Cisco Syslems. Inc
SIP is apeer-io-peer protocol where Internet endpoints (called user agents [UAs]) initiate
sessions similar to an H.323 peer. The UAs discover each other and agree on ai sessioni hat
thev would like to share. For locating session participants and other functions. SIP enables the
creation of an infrastructure of network hosts (called proxy servers) to which user agents can
send registrations, invitations to sessions, and other requests. SIP is an agile general-purpose
tool for creating, modifying, and terminating sessions, which works independently ot _
underlying transport protocols and without depending on the type of session that is being
established.
Unlike H.323, SIP uses ASCII text-based messages to communicate. Therefore, it allows for
easy troubleshooting by analyzing the signaling content.
" VoIP Call Legs 2-63
Signaling and Deployment
This subtopic explains the signaling functions and deployment methods ofSIP.
2-64
SiP Signaling
Determines the location of the target endpoint
Determines the media capabilities of the target endpoint
- SDP establishes the lowest common service level
Conferences use media parameters supported by all
Determines the availability ofthe target endpoint and informs why
the target was unavailable
Not reachable
Already connected to a call
No answer
Establishes a session between communicating endpoints
Including midcall changes (adding conference participants,
codec change)
Handles the iransferand termination of calls
SIP supports fiv cmethods ofestablishing and terminating multimedia communications which
result in the following capabilities:
Determines Ihe location ofthetarget endpoint: SIP supports address resolution, name
mapping, and call redirection.
Determines the media capabilities ofthetarget endpoint: SIP detcmiines the lowest
level ofcommon services between the endpoints through SDP. Conferences are established
using onlv [he media capabiliiies that can be supported by all endpoints.
Determines the availability ofthe target endpoint: If acall cannot be completed because
the target endpoint is unavailable. SIP detcmiines whether the called partv is connected to a
call alreadv or did not answer in the allotted number of rings. SIP then returns amessage
indicating whythe targetendpoint was unavailable.
Lstablishes a session between Ihe originating and target endpoints: If the call can be
completed. SIP establishes asession between the endpoints. SIP also supports midcall
changes, such as the addition of another endpoint to the conference or the chancing of a
media characteristic or codec.
Manages ihe transfer and termination ofcalls: SIP supports (he transfer ofcalls from
one endpoint to another. During acall transfer, SIP simply establishes asession between
the transferee and anew endpoint (specified by the transferring party) and terminates the
session between the transferee and the transferring party.
Implementing Cisco Voice Communications andQoS (CVOICE) v8 0
20l0Cisco Systems. Inc.
This subtopic describes the SIP architecture components.
SIP
SIP User Agents
(UAs)
SIP Proxy,
Registrar.
Location, and
Redirect Servers
.
SIP
UA Client - initiating party
UA Server - receiving party
RTP
SIP
Legacy PBX
SIP is apeer-to-peer protocol. The peers in asession are called UAs. AUA can function in one
of these two roles:
User agent client (UAC): Aclient application that initiates aSIP request
User agent server (UAS): Aserver application that contacts the user when aSIP invitation
isreceived and then returns aresponse on behalf ofthe user tothe invitation originator
Tv picallv. aUA can function as aUAC or aUAS during asession, but not both in the same
session. Whether the endpoint functions as a UAC or a UAS depends on the UA that initiated
the request: the UAC initiates and UAS terminates the session.
From an architectural standpoinU the physical components ofa SIP network are grouped into
these two categories:
Clients (endpoints)
Phone: An IP telephone acts as a UAS or UAC on asession-by-session basis.
Gateway: Agateway acts as aUAS or UAC and provides call control support. Like
in H.323. SIP gateways provide many services, the most common being atranslation
ftinction between SIP endpoints and other device types, such aspublic switched
telephone network (PSTN) destinations.
Servers: Registrar. Proxy, Location, andRedirection.
VoIP Call Legs 2-65
SIP Servers
This subtopic explains the server types ina SIP environment.
SIP Servers
Registrar server
- Accepts registration requests from users
Proxy server
Relays communications
Acts as client and server
Keeps no session state
- Transparent to end devices
Does not generate its own messages (except ACK and Cancel)
May add services (call forwarding, AAA, forking, and soon)
Redirect server.
Redirects callers to other servers
Rarely used to scale to large environments
Location server:
Maintains user whereabouts
Servers usually deployed as a single platform
The different server roles inthe SIP environment have these characteristics:
Registrar server: Receives requests from UACs for registration oftheir current location.
Registrar serv crs are often located near or even collocated with other network servers, most
often a location server.
Proxy server: An intermediate component that receives SIP requests from aclient and then
lonvards the requests onbehalf of the client lothe next SIP server in the network. The next
server can be another proxy server ora UAS. Proxy servers can provide functions such as
authentication, authorization, network access control, routing, reliable request
transmissions, and securitv.
Redirect server: Provides the client with infonnation about the next hop or hops that a
message should lake and then the client contacts the next-hop server or UAS directly.
When the redirect server sends a redirect message tothe client, theclient resends the
invitation tothe server identified in the redirection message. The client can be redirected
either to another network server orto the IIAS in the terminating endpoint.
Location server: Implements mechanisms loresolve addresses. These mechanisms can
include adatabase of registrations oraccess locommonly used resolution tools such as
Finger protocol, whois. Lightweight Directory Access Protocol (LDAP). or operating
system-dependent mechanisms. Aregistrar server can be modeled as one subcomponent of
alocation sen er: the registrar server is partly responsible for populating adatabase that is
associated with the location scner.
Note
SIP servers can interact with other application services, suchas LDAP servers, a database
application, oranXML application. These application services provide back-end services,
such as directory, authentication, and billing services.
2-66 Implementing Cisco Voice Communications andQoS (CVOICE) v8.0
This subtopic provides examples of SIP architecture.
Cisco Unified
Commimical ions
Manager
Gateway
SIPTrunk tram Carrier
Imere Re SIP Trunk
IPNetwork
Cisco Unified
Communications
Manager Express
Cisco Unified
Communi cations
Manager
Cisco Unified
Communi cations
Manager
Cisco Unified
Communications
Manager Express
Cisco Unified Communications implementations can deploy SIP on the following products:
Cisco Unified Communications Manager
Cisco Unified Communications Manager Business Edition
Cisco Smart Business Communication System
Cisco voice gateways
Cisco Unified IP phones running SIP firmware, which register on a Cisco Unified
Communications Manager or Cisco Unified Communications Manager Express
Cisco Unified IP phones running SIP firmware and connecting directly to an Internet
telephony service provider (ITSP)
SIP trunks to a carrier, and between corporate offices
SIP Call Flows
This topic describes the call setup and teardown flows in the three common scenarios: direct
call between SIP gateways, call over a proxy server, and call over a redirect server.
Direct Call Setup
CallingParty SlP Gateway
5 Ring Back Tone
2 Invite (SDP)
3 100 Trying
5 ISO Ringing
10 RTP Sir
11 BYE
IP
SIP Gateway
4 Ring
Called Party
Called Party
7 Answer Call
Signaling
This figuredepicts the direct call setup and teardown between two SIP gateways.
When a UAC recognizes ihe address of a terminating endpoint from cached infonnation. or has
the capacitv to resolve it by some internal mechanism, the UAC may iniliate direct (UAC-to-
UAS) call setup procedures. If a UAC recognizes the destination UAS. the client communicates
di-ecllv with the server. In situations in which the client is unable to establish a direct
relationship, the client solicits the assistance of a network server.
Direct call setup proceeds as follows:
Step 1 Endpoint initiates a call.
Step 2 The originating UACsends an invitation (INVITE) to the UASofthe recipient, fhe
message includes an endpoint description ofthe UAC and the SDP description ofthe
supported media parameters.
The UAS ofthe recipient responds to the INVl'l h message using the UK) Trying
message.
Ihe tenninating gateway sends the ringing signal lo the recipient telephone.
The recipient UAS informs the UAC aboul the ring signal with the Ringing
message.
The originating gateway sends the ringback tone to the caller telephone.
Ihe called telephone is taken off-hook.
If the UASofthe recipient determines that the call parameters are acceptable, it
Step 3
Step 4
Step 5
Step 6
Step 7
StepS
Step 9
responds positively to the originator UAC using the 200 OK message.
The originating UAC issues an acknowledgment (ACK) to the UAS.
Step 10 Atthis point, the UAC and UAS have all theinformation thatis required toestablish
Real-TimeTransport Protocol (RTP) sessions between them.
Step 11 One of participants temiinatcs thecall. Its UA sends theBYE message to the other
UA.
Step 12 The BYE message is confirmed by the 200 OK message.
SIP Call Setup Using Proxy Server
This subtopic describes the call setup and teardown in the scenario with a proxy server.
Calling Par
Proxy Ser
SIPGateway Called Parly
Ringing
-
IP
invite (SDP)
100Trying
180 Ringing
200 OK
ACK
The proxy server procedure is transparent to a UAC. Ihe proxy server intercepts and forwards
an invitation to the destination UAS on behalf of the originator.
A proxy sen er responds to the issues ofthe direct method by centralizing control and
management of call setup and providing a more dynamic and up-to-date address resolution
capability . The benefit to the UAC is that it does not need to learn the coordinates ofthe
destination UAS. yet it can still communicate with the destination UAS. The disadvantages of
this method include an increase in the signaling and the dependency on the proxy server. If the
proxy server fails, ihe LAC is incapable of establishing its own sessions.
Note Although the proxy server acls on behalf of a UA for call setup, the UAs establish RTP
sessions directly with each other
2-70 Implementing Cisco Voice Communications and QoS (CVOICE! v8.0 2010 Cisco Systems, Inc.
This subtopic describes thecall setup and teardown inthe scenario with a redirect server.
Calling Party SP Gateway P>xy Server SP Gateway CalledParty
Invite
100 Trying
1BORinging
200 OK
ACK
RTP Sfream
BYE
200 OK
Aredirect server is programmed todiscover a pathtothedestination. Instead of forwarding the
INVITE to the destination, the redirect server reports back to a UA with the destination
coordinates that the UA should try' next.
A redirect server implements many ofthe features ofthe proxy server. In the redirect server
scenario, fewer messages are exchanged than in the case ofthe proxy server. The UAC has a
heavier workload because it must initiate the subsequent invitation.
When a redirect server is used, the call setup procedure starts when the originating UAC sends
an INVITE to the redirect server, fhe redirect server, if required, consults the location server to
determine the path to the recipient and its IPaddress. The redirect server returnsa "moved"'
response totheoriginating UAC with the IPaddress obtained from the location server. The
originating UAC acknowledges theredirection andcontinues as described inthedirectcall
setup procedure.
VoIP Call Legs 2-71
SIP Addressing
I his topic describes SIP address formats, address registration, and address resolution.
SIP Address Types
* Address format uses Internet URLs
General form is name(domain
Address type
Fully qualified domain
namejFQDN)
E.164 (PSTN) address sip:14085551234@gateway.com;user=phone
Mixed format
Sip:14085551234;
password=changeme@10.1.1.1
SIP addresses use Internet URLs. Their general form is name^domain. An address in SIP is
defined in the syntax with "sip:" or "sips:"" (for secure SIP connections) as the URL type. The
URLs identity the originator, the current destination, the final recipient, and any contact party.
When two UAs communicate directly with each other, the current destination and final
recipient URLs arc the same. However, the current destination and the final recipient are
different if a proxy or redirect server is used.
To obtain the IP address of a SIP UAS or a network server, a UAC performs address resolution
ol a user identiiier. An address consists of an optional user II), a host description, and optional
parameters to qualify the address more precisely. The host description may be a domain name
or an IP address. A password is associated with the user ID, and a port number is associated
with the host description.
SIP Addressing Variants Example
'fhe figure provides c\amples of SIP addresses.
In the example "sip: 14085551234 c/gateway.com: user=phone." the *'user=phone" parameter is
required to indicate that the user part ofthe address is a telephone number. Without the
"user=phone" parameter, the user ID is taken literally as a numeric string. The "14085559876"
in the URL "sip: 14085559876 (7,10.1.1.1" is an example of a numeric user ID. In the same
example, the password "changeme" is defined for the user.
2-72 Implementing Cisco Voice Communications and QoS (CVOICE) v8.0 >2010 Cisco Systems, Inc.
This subtopic describes the process of address registration in SIP.
Although SIP servers aretypically
collocated, the UAclients:
Firstregisteron the registrarserver
Place calls over Ihe proxy server
ASIP address is acquired in several ways: by interacting with auser, by caching information
from an earlier session, or by interacting with anetwork server. The network servers must
recognize the endpoints in the network. This knowledge is abstracted to reside in alocation
server and is dynamically acquired by its registrar server.
To contribute to this dynamic knowledge, an endpoint registers its user addresses with a
registrar server The figure illustrates avoice register mode request to aregistrar server. When
the registration is complete, the information about the UAC is entered into the location
database, and the proxy server will be able to provide the endpoint address when other
endpoints wish to contact it.
i 2010 Cisco Systems, inc.
VoIP Call Legs 2-73
Address Resolution
This subtopic described the process ofaddress resolution.
Address Resolution
Where is tlie name
or phone number"?
When SIP proxy receives INVITE message
requesting call to an endpoint, it:
Authenticates the caller
Looks up the locationdalabase to resolve the
endpoint address FQDN, E.164, or mixedto its
current IP address
When an endpoint attempts to communicate, it must resolve the IP address ofthe destination
endpoint ihat is based on its address in the fullv qualified domain name (FQDN) F164 or
mixed address format. To resolve an address, aVA uses avariety ofinternal mechanisms such
as alocal host table and DNS lookup, or. more commonly, it leaves that responsibility to the
proxy serv er. Ihe proxy server uses any ofthe tools available to aUA orinteracts with the
ocation server. In the ligure. the SIP proxy server interacts with alocation server to derive the
location ol the end dev ice in question. Once the IP address ofthe destination endpoint is
established, the SIP proxy fonvards the call to the destination device, or the redirect server
responds to the initiating endpoint with the address ofthe destination party.
2-74 implementing Cisco Voice Communications and QoS (CVOICE) v8 0
Codecs in SIP
This topic describes codec negotiation in SIP.
Session Description Protocol
SDP describes session parameters in SIP.
SDP carries:
- The type of media (video, audio, etc.)
The transport protocol (RTP/UDP/IP, H.320, etc.)
- The format of the media (codecs)
A list of media formats can be offered:
- All listed formats may be used in the session.
- The first format is the default format.
Exchanges codecs at different stages in call setup :
- Delayed Offer: 200 OK and ACK
Early Offer: Invite and 200 OK
- Early media: 183 Session Progress, 180 Ringing, Pre-Ack
SIP leverages a number of other standards-based protocols to provide a large set of features
based on relatively simple mechanisms. One ofthe relevant protocols is the SDP.
The SDP is an lETF-based format for describing streaming media initialization parameters in
an ASCII string. SDP is intended for describing multimedia communicationsessions for the
purposes of session announcement, session invitation, and parameter negotiation. SDP does not
deliver media itself but is used for negotiation between endpoints of media type, format, and all
associated properties. The set of properties and parameters are often called a session profile.
SDP is designed to be extensible to support new media types and formats.
SIP leverages SDP to negotiate the type of media (audio, video), the transport protocol (RTP or
User Datagram Protocol [UDP] ports), and the format of media (audio and video codecs). The
initiating endpoint can provide a list of capabilities, while the first offer is the default (highest-
priority) proposal. The destination endpoint selects an offer that matches its capabilities and
keeps the complete list of common capabilities in case the capabilities should be changed
midcall.
SIP uses the Offer/Answer model for establishing SIP sessions. An Offer is contained in the
SDP fields that arc sent in the body of a SIP message. The Offer defines the media
characteristics that are supported by the device (media streams, codecs, directional attributes. IP
address, and ports to use). The device receiving the Offer sends an Answer in the SDP fields of
its SIP response, with its corresponding matching media streams and codec, whether accepted
or not. and the IP address and port on which it wants to receive the media streams.
SDP Examples
This subtopic provides examples of SDP messages.
Examples
Example 1. Audio, RTP/49100. G 711 mu-law)
o=b]oe +1-201-S5S-1212 IN IP*
hos tl. Cisco, cm
scExample-
t=0 0
C=IK IP4 L 92. 168. 1.1
m=audio 4910C RTP/AVP 0
Example 2. Audio. RTP/3456, G .729 most
preferred. G.711 mu-law second choice. G 711
a-law third choice)
v=0
o=a mith 1301555676S IN IP4 ciseo.com
s=Exampl e2
t = 0 0
C=IN IP4 10 234 . 1 . 1
m=a dio 3456 RTP/AVP IB 0 B
SDP content varies depending
on the message type
Version v=Q
Origin o=<jsernamei ^session id>
*versiorii ^networktype>
=aOdress type> <afldressi
Session
name
s"sesslon nams>
Times t=<start lime> <stop time?
Connection
data
c=<ne toorfctypes <ad drass type>
^connection addras3>
Media m=emedia><port> ^transports
<media format list*
Audio
Video
Profile
(AVPj
codes
0:G.711 mu-Law
B.G711 a-law
3.GSM codec
18:G.729
This figure presents two SDP examples and the table explains the parameters that are used in
these two examples as follows:
Version: Protocol version
Origin: Describes the sender ofthe message, may include one or more of these parameters:
username. session ID. address tvpe. and the address value.
limes: Optionallv defines the session start and end limestamps. The values are not set
when a call setup is signaled.
Connection data: Provides the parameters for media endpoint termination: network type
("IN" is defined as "Internet." and other types may ho added), address type (IPv4/v6), and
the connection address {IP address)
Media: Specifics the media tvpe (audio/video), the UDP transport port, and one or more
media formats. Fxamples of Audio Video Profile (AVP) codes arc: 0(G.7I I mu-law). 8
(G.711 a-law). 3 (Global System for Mobile Communications [GSM| codec), 18 (G.729).
The list is ordered according to the priority.
SDP content varies depending on the message type.
2-76 Implementing Cisco Voice Communications and QoS (CVOICE] v8 0 2010 Cisco Systems, Inc
Delayed Offer
This subtopic describes how SDP negotiates codecs using SIP Delayed OtTer.
Delayed Offer
Calling Party SIP Gateway SP Gateway
S Ringback Tone
2 Invite
3.100 Trying
5 180Ringng
8. 200OK [SDP: media offer)
9 ACK (SDP media answer)
10 RTP Stream
11. BYE
12 200 OK
Called Party
There are two ways to exchange the SDP Offer and Answer messages. These methods are
commonly known as Delayed Offer and Early Offer, and support for both methods by user
agent client/servers is a mandatory' requirement ofthe SIP specification. In the simplest terms,
an initial SIP Invite that is sent with SDP in the message body defines an Early Offer, whereas
an initial SIP Invite without SDP in the message body defines a Delayed Offer.
In a Delayed Offer, the session initiator does not send its capabilities in the initial Invite but
waits for the called device to send its capabilities first (for example, the list of codecs that arc
supported by the called device, thus allowing the calling device to choose the codec to be used
for the session).
The Delayed Offer is recommended for SIP trunks because it enables the ITSPs to provide their
capabilities first. The Cisco Unified Communications Manager allows the administrator to
select the offer method. Cisco gateways support both methods but originating gateways default
to Larly Offer.
>2010 Cisco Systems, Inc. VoIP Call Legs 2-77
Early Offer
2-78
This subtopic describes how SDP negotiates codecs using SIP Early Offer.
CallingParty siPGaleway
6 Ringback Tone
Default on Cisco
gateways iSDP
in Invite nessagel
IP
2 Invite (SDP media offer]
8 200 OK (SDP media answer)
9 ACK
10 RTP Stream
11 BYE
SIPGateway Called Party
7 Answer Call
In an Karlv Offer, the session initiator (calling device) sends its capabiliiies (including
sapported codecs) in the SDP contained in the initial Invite, fhis method allows the called
device to choose its preferred codec for the session. Parly Offer is the default method that is
used by a Cisco voice gatewav acting as the originating gateway.
Implementing Cisco Voice Communications and QoS (CVOICE) vB.O ) 2010 Cisco Systems, Inc
Early Media
This subtopic describes the SIP Early Mediafeature.
Early Media
183 Session Progress Option
Calling Party SIP Gateway
100 Trying
SP Gateway
Earty Media allows
trie sending ol media
from the called party
or an application
server to the caller,
pnorto tne cal being
accepted Early
media is usually sent
from the PSTN and
carries nnging lones
or announcements
183 Session Progress (SDP Media Offer)
Pre-ACK (SDR Media Response)
180 Rmgng
RTP aream
Called Party
SIP Early Media was originally defined in RFC 3960 as a facility for PSTNinterworking. Early-
Media allows the sending of media from the called party or an application server to the caller.
even before the call is accepted. The most common reasons for using Early Media include the
following:
The called device might want to establish an Earty Media RTP path lo reduce the effects of
audio cut-throughdelay (clipping) for calls experiencing long signaling delays or to
provide a network-based voice message to the caller.
The calling device might want to establish an Early Media RTP path to access a dual tone
multifrequency (DTMF) or voice-driven interactive voice response (IVR) system.
Cisco gateways support Early Media for both Early Offer and Delayed Offer calls.
If no media is available for streaming at this early stage, the early media channels carry silence.
Voice activity detection (VAD). if negotiated, would in that case prevent bandwidth
consumption by dropping silence packets.
With Early Offer (default on Cisco gateways), the SDP offer is carried in the INVITE message.
In Early Media with Delayed Offer, both messages can transport the initial SDP offer: 183
Session Progress response or 180 Ringing response. 183 Session Progress is stipulated by the
IETF and is more common. The 183 Session Progress response indicates that infonnation about
the call state is present in the message body media information. The SDP media response is
exchanged in an additional pre-ACK message, after which the endpoints can establish the RTP
streams.
Early Media (Cont.:
180 Ringing Option
Calling Party
SIP Gateway
JETF draft allows
olher-than-183
messages to cariy
SDP Some
implementations use
180 Cisco gateways
accept the 180
methcd by default (in
addition to 1831 This
method can he
disabled on Cisco
galeways
100 Trying
180 Ringing (SDP Media Offer)
Pre-ACK (SDP Media Response)
RTP Stream
200OK
ACK
SIP Gateway Called Party
IP
To facilitate Earlv Media v\ ith Delayed Offer, the IETEdraft allows the use of other messages
than the 183 Session Progress response. Some implementations use the 180 Ringing response
to send the initial SDP media offer. The 180 Ringing message is a provisional or informational
response that is used to indicate that the INVI IT' message has been received by the user agent
and that alerting is taking place. Cisco gateways support both 180 and 183 methods to negotiate
Early Media.
Cisco gatewav s. bv default, process a 180 Ringing response with SDP in the same manner as a
183 Session Progress response: thai is. ihe SDP is assumed to be an indication that the far end
would send carlv media. This behav ior can be changed so that a gateway ignores the presence
or absence of SDP in 180 messages, and as a result. Ireat all 180 messages in a uniform manner.
Implementing Cisco Voice Communications and QoS (CVOICE) v8 i 2010 Cisco Systems, Inc
a^tf
Configuring Basic SIP
Thistopic describes howloconfigure basicSIPfunctionality on voice gateways.
Basic SIP Configuration Overview
Configure SIP User Agent (UA):
- Authentication
- SIP servers
Configure SIP-related dial peer parameters:
- Session protocol
- Session target
A SIPconfiguration consists of twoparts: the SIP UAand the VoIPdial peersthat select SIPas
the session protocol.
The basic UAC configuration includes the following:
Authentication parameters: username and password
SIP servers (registrar and proxy)
SIP dial peers have these two basic parameters that are specific to SIP:
Session protocol
Session target
) 2010 Cisco Systems, inc. VoIP Call Legs 2-81
User Agent Configuration
This subtopic explains the basic configurationof a SIP user agent.
2-82
User Agent Configurate
route r(config)#
sip-ua
Enters SIP UA configuration mode
router(config-sip-ua)#
registrar { dhcp [index] registrar-address [.-port]
* Register E 164 numbers on behalf of analog phones (FXS), IP phone
virtual voice ports (EFXS), and SCCP phones with an external SIP
proxy or SIP registrar
Up to six configurable registrars, can be obtained via DHCP
router (e on fig -sip-ua) # _____^
authentication username username password [0)7] password I
Enables SIP digest authentication
Only one username can be configured globally in SIPUA
To configure SIP user agent parameters, enter SIP UA conliguralion mode using the sip-ua
command.
The registrar command enables the gatewav to register L.I 64 numbers on behalf of analog
telephone voice ports (foreign Exchange Station [FXS]), IP phone virtual voice ports
(enhanced FXS [FFXS]). and Skinny Client Control Protocol (SCCP) phones with an external
SiPproxv or SIP registrar. It defines the IP address of the registrar server.
1 le full command svntax is registrar [dhcp | \registrar-index] registrar-server-
address\:port]\ [expires seconds] [random-contact] [refresh-ratio ratio-percentage J [scheme
{sip | sips!] [tcp] [type] [secondary]
fhe registrar address can be obtained via DHCP, ihe registrar-index option allows the
configuration of up to six registrars that can be used concurrently for redundancy and load-
balancing purposes. Further options allow the use of Secure SIP, TCP transport, and the
definition of a registrar pair (primary and secondary) instead of multiple indexed servers.
To enable useniamc-based message digest authentication ofthe user agent, configure the
authentication username command in UA configuration mode. This command delines the
username and password that the gateway uses to authenticate on the registrar server.
implementing Cisco Voice Communications and QoS (CVOICE) v8 0 2010 Cisco Systems, Inc
This subtopic describes the SIP-related parameters in basic SIP configuration.
sip-server {dns:boat-name \ ipv4:ipv4-address |
ipv6:[ipv6-addrass][:port-num]}
Defines a SIP server to be referenced in dial peers
router(config-dial-peer)#
session target sip-server
References the configured SIP server instead of its IP address
session protocol sipv2
Defines SIPv2 as session protocol
The sip-server command is a time-saving method. If you use this command, you can also use
the session target sip-server command on each dial peer instead of repeatedly entering the SIP
server interface address for each dial peer. Configuring a SIP server as a session target is useful
if the gateway acts as a UAC and makes calls over a SIP proxy. Multiple dial peers can
reference the same proxy server.
The session protocol sip\2 command enables a dial peer to use SIP version 2 as the signaling
protocol for a particular dial peer. The default value is H.323.
Basic SIP Configuration Example
This subtopic presents a basic configurationexample of SIP functionalityon a voice gateway,
Cisco Untied
Communications Manager
10 1115
192 168 1 100 ,
jthentication username JDoa password
registrar 10.1.1.15
sip-seiver 10.1.1.15
I
session target ipv4:192.16B.1.100
session protocol aipv2
SIP ITSP
I ie figure shows a voice galewav that communicates via SIP with two external SIP servers:
Cisco Unified Communications Manager and a SIP service that is operated by an ITSP.
fie Cisco I inified Communications Manager (with IP address 10.I.I. 15) includes two
collocated components: SIP registrar and SIP proxy. The SIP UA refers to the registrar
component using the registrar command and references the proxy component using the sip-
server command. The I'A configured on the gateway uses the dial peer 2001 to match the
destination patterns 2... and connect to the SIP proxy running on the Cisco Unified
Communications Manager (session target sip-server command points lo the address set with
sip-server command in sip-ua mode). The gateway will register on the Communications
Manager using the credentials that are defined in the authentication command.
For all other destinations that use the prefix 9 to represent the outside world, the dial peer 2002
points via SIP version 2 to the ITSP SIP proxv.
Implementing Cisco Voice Communications and QoS (CVOICE) v8.0 2010 Cisco Systems. Inc
Configuring SIP ISDN Support
This topicdescribes howto configure SIP ISDN support.
SIP ISDN Support Configuration Overview
SIP features for ISDN support:
- ISDN calling name display
Blocking caller IDwhen privacy exists
- Substituting the calling number for the display name,
if display name unavailable
SIPcanbe configured for variousISDN features. The most relevant ISDN functions that apply
to most situations are as follows:
ISDN calling name display
Blockingcaller ID when privacy exists.
Substitutingthe calling number for the display name, if the display name is unavailable
Calling Name Display
This subtopic describes the ISDN calling name display feature.
Calling Na
Enables SIP
for calls that
me Display
P phones to display caller-name identification
originate on an ISDN network
i, cs -en Incoming Call
y -~^ -T-..S- PRIORI ^L
Called SP SiP Gateway SIPGaleway ' Caller
Plane
In ISDN networks, caller ID (sometimes called calling line ID [CI.IDJ or incoming calling line
icentification [ICLIDJ) is a service thai is offered by a central office (CO) to supply calling
partv infonnation to subscribers. Caller IDallows the calling parly number and name to appear
on a device such as a telephone displav.
ISDN messages signal call control and are composed of infomialion elements that specify
screening and presentation indicators. ISDN messages and their information elements are
passed in Generic Transparency Descriptor ((il D) format. GTDenables transport of signaling
data in a standard fonnat across network components and applications. The standard fonnat
enables other dev ices to scan and interpret the data. The SIP network extracts the calling name
from the G'l'D fonnat and sends the calling name infonnation to the SIP endpoint.
Implementing Cisco Voice Communicalions and QoS (CVOICE] vB.O '2010 Cisco Systems. Inc
This subtopic explains how toconfigure the SIP ISDN calling name display feature.
router(conf-vol-serv)#
I signaling forward {none | unconditional}
Specifies whether or not the originatinggateway forwards the signaling
payload to the terminating gateway
None-do not pass the signaling payload to terminating gateway
- Forward the signaling payload unconditionally
- Configured in voice service voip configuration mode
router(config-if)#
isdn flupp-service name calling
Sets the calling-name display parameters sent from an ISDN serial interface
Configured in serial interface created on a channelized E1 or channelized
T1 controller
Whenan ISDNsubscriber placesa call to a SIPendpoint,the subscribercallingnumberis by
default supplied tothe SIP endpoint andappears onthedisplay when thecall comes in. The
calling name is typically not forwarded bydefault. Twocommands are needed toenable the
calling name display:
signaling forward: Thiscommand is issued in thevoice service VoIP configuration mode.
It specifies whethertheoriginating gateway forwards the signalingpayloadto the
terminating gateway. Keywords are as follows:
none: Prevents the gateway frompassing the signaling payload to die terminating
gateway.
unconditional: Forwards the signaling payload received in the originating gateway
to the terminating gateway, even if the attached external routeserver has modified
the CiID payload.
isdn supp-scrvice name calling: lhis command is issued in the configuration mode ofthe
serial interface that is created on a channelized FIAT controller. The command sets the
calling-name display parameters that are sent out an ISDNserial interface.
) 2010 Cisco Systems. Inc. VoIP Call Legs
Calling Name Display Configuration
This subtopic prov ides a configurationexample for the calling name display feature.
ling Name Dm
Called SIP SIP Gateway
Phone
Incoming Call
SIPGateway
ignaling forward unconditional
iterface serial 1/0:23
isdn supp-service name calling
The figure shows how to configure the calling name display feature on a voice gateway that is
connected to the PSTN via a Tl channelized controller using ISDN PRI signaling. The serial
interfaceand the voice service VoIP are configured to unconditionally forward the signaling
information that results in the calling name being displayed on the SIP endpoint when a call
arrives.
Implementing Cisco Voice Communications and QoS (CVOICE] v8.0 2010 Cisco Systems, Inc.
Blocking and Substituting Caller ID
This subtopic explains how toblock orsubstitute the caller ID.
Blocking and Substituting Caller \D
ISDN has a private setting to protect caller ID.
SIP does not hide the private information:
- Just sets a field to mark as private and not to display it
~ Data still viewable in the SIP message requests
Blockoption deletes the caller ID information fromthe SIP message
requests so that it cannot be read on the network.
With substitution, ifthere is no Display Name field but there is a number,
the number is copied intothe Display Name fieldand presented on the
displayof the recipient.
Called SP SIP Gateway
Phone
SIP Gateway
Incoming Call
ThecallerIDinformation is private information. InISDN, there is a private setting that canbe
set to protect this information. However, when SIPgetsthecallerIDinformation, it doesnot
hide the private infonnation. Rather, itjust setsa field to reflect that it is private and not to
display it on a caller IDdisplay. Butthe data is still viewable inthe SIPmessagerequests.
The block option allows thegateway todelete thecallerIDinformation from theSIPmessage
requests so that it cannot be read on the network.
Thesubstitution option is helpful if there is no Display Name field but there is a number and
the presentation is not prohibited. Inthatcase it copies the number into the Display Name field,
so that the number is displayedon the caller IDdisplay ofthe recipient. The Ciscogateway
omits the Display Name field if no display information is received.
VoIP Call Legs 2-89
Blocking and Substituting Caller ID Commands
This subtopic presents the commands that areneeded toconfigure CLID blocking and
substituting CLID for the display name.
2-90
Blocking and Substituting Cai
Commands
router(config-dial-peerI#
did strip pi-restrict
Block caller ID information when privacy exists
- Available in voice service voip and dial peer configuration modes
router(cont-vol-serv)#
did substitute name
Substitutes the calling number for the display name when the display name
is unavailable
Available in voice service voip and dial peer configuration modes
This figure presents thecommand sv nla\ fortwofeatures. You canenable both, eitherglobally
or on specific dial peers, I he global setting is configured in the voice service VoIP
configuration mode. The dial peer setting is configured in the dial peer configuration mode.
Issue the did strip pi-restrict command lo enable CLID blocking when privacy exists.
Issue the did substitute name command lo enable substitutionof CLID for the displav
name v\hen the display name is unavailable.
Implementing Cisco Voice Communications and QoS (CVOICE) v8.0 2010 Cisco Systems. Inc
Blocking and Substituting Caller ID Configuration
This subtopic presents an example ofblocking CLID with privacy and substituting CLID for
thedisplav name if thedisplay name is unavailable.
Blocking and Substituting Caller
Configuration
Incoming Call
voice service voip
did substituta name
clid strip pi-restrict
line figure showsan examplewithtwo features enabled:
The feature to substitute CLIDfor the displayname whenthe displayname is unavailable
is enabled in the voice service VoIP configuration mode and applies to all calls processed
by the gateway.
The feature to blockCLIDwhenprivacyexists is enabledin the dial peer configuration
modeand appliesto the calls forwarded usingthis specificVoIPdial peer setting.
i 2010 Cisco Syslems, Inc. VoIP Call Legs
Configuring SIP SRTP Support
2-92
This topicdescribes how to configure securesignaling and securemedia inthe SIP
env ironmcnt.
SIP SRTP Support Overvie
Two independently configured security areas:
- Signaling
SIP Secure protected using TLS
Media
Secure RTP (AES encryption, HMAC-SHA1 authentication)
ESim&) srtp
On On
Off On
On Off
Off Off
Signaling and media are secure.
Signaling is insecure or secured with other methods.
Media is secure wltt Cisco IOS Release 12.4(22)T and
later. Media falls back to RTP or fails m earlier versions.
Media insecure (RTP-only)
Signaling and media insecure
SIP offers two methods to secure voice communications:
SIP secure (SIPS) offers signaling authentication and encryption using the Transport l.aver
Security (TLS) protocol. When TLS is used, the cryptographic parameters that are required
to successful!} negotiate Secure Real-Time Transport Protocol (SRTP) relv on the
cryptographic attribute in the SDP, To ensure the integrity of cryptographic parameters
across a network. SRTP uses the SIPS schema.
SRTP offers media authentication (Hashed Message Authentication Code-Secure Hash
Algorithm I [IIMAC-SHA-L|) and encryption (Advanced Encryption Standard [AKSJ) to
secure the media flow between two SIP endpoints. lypically, SRTP is used in combination
with SIPS, although SIPS is no longer required for SRTP in Cisco IOS Release 12.4(22)1
and later. Calls established with SIP (and not SIPS) can still successfully negotiate SRTP.
In such cases, the signaling should be protected using a different protocol, such as IPscc.
fhe table in the figure shows various combinations ofthe SIPS and SRI T settings. The second
combination (SIPSdisabled. SRI Penabled) results in varying behavior, depending on the
Cisco IOS release. With Cisco IOS Release 12.4(22)Tand later, the signaling is in elearlexl and
the media is encrvpted. With earlier releases, the calls either fall back to RTPor fail, depending
on the securcrtp fallback command.
Implementing Cisco Voice Communications and QoS (CVOICE) v8 0 i 2010 Cisco Systems, Inc
This subtopic explains the commands that are needed toconfigure SIPS and SRTP.
router(conf-dial-peer)#
voice-clase aip
Enters dial peer SIP configuration mode
router(conf-voi-eer)#
sip
Enters SIP configuration mode, in:
- Voice service VoIP configuration mode, or
- Dial peer voice dass SIP configuration mode
router(conf-ser-flip)*
url sips
Enables SIPS by generating URLs in SIPS format for VoIPcalls
Available globally or in dial peer configuration mode
SIPS functionality was introduced inCiscoIOS Release 12.4(15)1. Youcanconfigure secure
signaling onbotha global level (in SIPmode) andon anindividual dial peerbasis, lo
configure SIPS globally, youmust firstenterthevoice service VoIP configuration mode (voice
service voip command) and then the SIPconfiguration mode(sip command). To enable SIPS.
issue the url sips command.
Thedial peersettingoverwrites theglobal setting, which is useful when disabling SIPS on
selecteddial peers whenSIPSis enabledglobally. To configureSIPSfor a dial peer, you must
first enterthedial peervoice classSIPconfiguration mode(voice-class sip command, followed
by the sip command) from the dial peer configuration mode. To enable SIPS invoice-class SIP.
issue the url sips command.
VoIP Call Legs
SRTP Global and Dial Peer Commands
"I his subtopic describes the commands that are needed for configuring SRTP
router(conf-dial-peerI #
voice-class sip
Enlers dial peer SIP configuration mode
router{conf-vol-ser)#
securertp
* Configures secure RTP media, in.
Voce service VoIP configuration mode, or
Dal peer voce class SIP configuration mode
router(conf-vol-ser 11
router(conf-dial-peer]#
Iaecurertp fallback
Enables fallback to RTP calls in case secure RTP calls fail due to lack of support
from anendpoint
Available globally or in dial peer configuration mode
SRTP was introduced in Cisco IOS Release 12.4(15)1. You can configure the secure media
transport on both a global level (in SIP configuration mode) and on an individual dial peer
basis. The dial peer setting overwrites the global setting. To configure SRTP for a dial peer,
vou musl first enter the dial peer voice class SIP configuration mode from the dial peer
configuration mode. To configure SRTP globally, you must first enter the voice service VoIP
configuration mode (voice service voip command) and then the SIP configuration mode (sip
command). To configure SRTP in either mode, issue the securertp command. If you configure
the gatewav for SRTP (globally or on an individual dial peer) and end-to-end TLS, an outgoing
1NV1TL message has cryptographic parameters in the SDP.
You can also configure the gateway (or dial peer) either to fall back to (nonsecure) RTP or to
reject (fail) the call in cases where an endpoint does not support SKIP, "fhis behavior is
configured v\ ith the securrtp fallback command, issued either in SIP or dial-peer voice-class
sip configuration mode. If you use the srtp fallback command and the called endpoint does not
support SRTP (offer is rejected with a 4xx class error response), the gateway sends an RIP
offer SDP in a new INVITh request.
Implementing Cisco Voice Communications and QoS (CVOICE) v8 0 ) 2010 Cisco Systems. Inc
This subtopic presents a configuration example for SIPS and SRTP.
1001 SIPGWMvay
voice nervine voip
sip
url sips
sec rertp
see rertp fallback
dial-peer v ice 1 vo IP
dest natio -pattern 2.. .
sess on protocol Sipv2
session target ipv4 10.2.1.1
session target ipv4:lD.1.1.1
voice-class sip
eecurertp
securortp fallback
sip
url sips
The figure shows die configurationof two voice gateways that are configured for SIPS and
SRTP. The gateway on the left has the settings configured globally, while the right gateway is
configured on a specific dial peer. Both support fallback to RTP in case SRTP is not supported
by the other endpoint.
Customizing SIP Gateways
This topic describes the tuning options in SIP.
2-96
SIP Gateways Tuning Overview
Tuning Options:
* SIP transport
Global SIP
Dial peer
User agent
* Source IP address
Global SIP
SIP Timers
User agent
* Early media
Handling of 180 Ringing responses with SDP
flic most common SIP customization tasks include the following:
Defining the session transport protocol: I'CP. I'CP-TLS. or UDP. fhis setting can be
applied in global SIP. dial peer, or UA configuration mode.
Selecting a source IP address bv binding the gateway functionality to a network interface.
This option is available onlv in global SIP configuration mode.
Tuning SIP timers: I hesc parameters are tunable in the UA configuration mode.
Defining which call treatment, earlv media, or local ringback. is provided for 180 Ringing
responses with SDP.
Implementing Cisco Voice Communications and QoS (CVOICE) v8 0 '2010 Cisco Systems, Inc
SIP Transport
This subtopic describes how to configure SIP transport protocol
SIP Transport
router(conf-voi-eer)#
session transport {system | top tls | udp}
Defines SIP transport globally or for dial peer
- In SIP mode without system option
- System option (in dial peer mode) refers to SIP global mode
Applies to outgoing signaling
Default: UDP
router(conflg-slp-ua)#
transport {top tls | udp}
Enables the UAto receive signaling messages for inbound calls
overTCP, TCP TLS, or UDP. Uses port 5060.
Bydefault, all three transports are enabled.
The configuration of SIP session transport refers to two aspects of signaling:
Outbound signaling: Default is UDP. The transport for outgoing SIP messages can be
configured globally, in SIP configuration mode, and in the dial peer configuration mode.
The system option in the dial peer mode applies the global option to a specific dial peer and
is used as a time saver. Instead of configuring a non-UDP option repeatedly for each dial
peer, you can configure the global setting and apply it to the required dial peers.
Inbound signaling: This option is configured in the SIP UA configuration mode. It
specifies the transport methods accepted for receiving inbound calls. The default is lo
accept all three transports: UDP. TCP, and TCP TLS, on port 5060.
VoIP Call Legs 2-97
SIP Source IP Address and UA Timers
This subtopic describes how to configure the interface binding feature and tune SIP timers.
router(conf-vol-ser)#
bind {control media all} source-interface interface-id
[ ipv4- address ipv4-addreBs | ipv6-address ipvfi-address]
Binds the source address for signaling and media packets to the
address of a specific interface
Configures various timers for the SIP UA
fhe interface binding feature sets the IP address for outgoing SIP-related traffic. To configure
the interface binding feature, issue die bind command in the global SIP configuration mode.
You have the option to hind either signaling, media, or both, using the control, media, and all
kevwords. The command points to an interface and specifics its IPv4 or IPv6 address that
should be used as the source IP address for outgoing traflic.
To tune SIP timers. \ou must enter the SIP UA configuration mode.
2-98 Implementing Cisco Voice Communicalions and QoS (CVOICE) vS 0 2010 Cisco Systems, Inc.
SIP UA Timers
This subtopicexplainssome SIP UAtimers.
SIP UA Timers
Timer I Description
Default
connect Time (in milieeconds) to wait for a 200 response to an
ACK request
500
disconnect Time (in miliseconds) to wait for a 200 response to a BYE
request
500
expires Time {inmiliseconds) for which an INVTTE request is valid 180000
hold Time to wait during hold before disconnecting (in minutes) 2880
notify Time to wa9 before NOTIFY retransmission 500
refer Time to wait before REFER retransmission. Refer request
is sent by the originating gateway to the receiving gateway
and initiates call forward and call transfer capacities
500
register Time to wa before REGISTER retransmission 500
trying Time (in miliseconds) lo wait for a 100 response loan
INVITE request
500
The default values of SIP timers work well in most environments and should not be changed
unless the administrator identifies a specific requirement. These timers can be set in the SIP UA
configuration mode:
Connect: Time (in milliseconds) to wait for a 200 response to an ACK request. Range is
from 100 to 1000. The default is 500.
Disconnect: Time (in milliseconds) to wait for a 200 response to a BYE request. Range is
Expires: Time (in milliseconds) for which an INVITE request is valid. Range is from
60000 to 300000. The default is 180000.
Hold: Time (in minutes) to wail before disconnecting a held call by sending a BYE
request. Range is from 15 to 2880 minutes. The default is 2880.
Notify: lime (in milliseconds) to wait before retransmitting a Notify message. Range is
Refer: Time (in milliseconds) to wait before retransmitting a Refer request. Range is from
100 to 1000. The default is 500.
Register: Time (in milliseconds) to wait before retransmitting a Register request. Range is
Trying: lime (in milliseconds) to wait for a 100 response to an INVITE request. Range is
>2010 Cisco Systems. Inc. VoIP Call Legs 2-99
SIP Early Media
fhis subtopic describes howto disable SIP Earlv Media for 180 Ringing messages.
router{config-sip-ua)#
disable-early-media 180
Disables early media cut-through treatment for SIP 180 Ringing
messages with SDP
Configured in SIP user agent mode
Does not affect treatment of SIP 183 Session Progress messages
* Early media enabled by default for SIP 183 Session Progress
messages
180 response with SOP Enabled (default)
180 response with SDP Disabled
180 response without SDP Not affected
183 response with SDP Not affected (default enabled)
Earty media cut-through
Local ringback
Local ringback
Early media cut-through
I he SIP Enhanced 180 Provisional Response 1landhng feature provides the ability to enable or
disable earlv media cut-through on Cisco IOS gatewavs for SIP 180 response messages. This
leature allows vou to speeifv whether 180 messages with SDP are handled in the same way as
183 responses with SDP, The 180 Ringing message is a prov isional or informational response
that is used to indicate that the INVITE message has been received by the user agent and that
alerting is taking place, fhe 183 Session Progress response indicates that information about the
call state is present in the message body media information. Both 180 and 183 messages ma\
contain SDP. which allow an earlv media session to be established prior to the call being
arswered.
Bv default. Cisco gatewavs handle a 180 Ringing response with SDP in the same manner as a
IS3 Session Progress response: that is. the SDP is assumed to be an indication that the far end
would send early media. Cisco gatewavs handle a 180 response without SDP by providing local
ringback. rather than early media cut-through, fhis feature provides the capability to ignore the
presence or absence of SDP in 180 messages, and as a result, treat all 180 messages in a
uniform manner. The disahle-early-media 18(1 command allows specifying which call
treatment, earlv media, or local ringback is provided for 180 responses with SDP.
2-100 Implementing Cisco Voice Communications and QoS (CVOICE) vB.O i 2010 Cisco Systems, Inc.
This subtopic presents a gateway-to-gateway configuration example.
LoopbaekC
irj01 10 111
voice service voip
Hip
session transport top
bind all source-interface
lookback 0 ipv4-address 10.1.1.1
I
session protocol Bipv2
BBBsion target ipv4:10.2.1.1
sip-ua
disabla-early-madia ISO
LoopbaekO
1021 1 2001
voice
aip
bind all source-interface
loopbaek 0 ipv4-address 10.2.1.1
1
session protocol Sipv2
session target ipv4slO.1.1.1
1
sip-ua
disabla-early-madia ISO
ervic =ip
This example shows two voice gateways that signal calls via SIP. Both gateways source the
signaling and media traffic from the IP addresses configured on their respective loopback 0
interfaces. Both gateways use TCP as the transport protocol for outbound signaling. The dial
peer I on Rl refers to the system setting that is configured in the SIP mode. The dial peer I on
R2 has the transport that is configured in its dial peer settings. If dial peer 1on Rl would not
have the session transport system command, it would signal calls to R2 using UDP transport.
R2 would accept that traffic, because the supported transports for inbound signaling are
configured in sip-ua mode and, by default, include all three options: UDP, TCP, and TCP TLS.
SIP 180 Ringing responses carrying SDP media offers are ignored.
) 2010 Cisco Systems, Inc. VoIP Call Legs 2-10!
UA Example
2-102
This subtopic prov ides a configurationand tuning example for a SIP UA.
!
192 16S 1.100 ,
SIP ITSP
bind all source-interface loopbackO ipv4-address 10.1.1.1
authentication username JDue password
registrar 10.1.1.15 expires 3600
sip-server 10.1.1.35
timers connect 1000
timers register 300
session target ipv4:192.1SB.1.100
The example in the figure shows some customization commands on a voice gateway that
communicates \ ia SIP with an external SIP servor operated by an ITSP.
All outgoing SIP and media communications are sourced from the Eoopback 0 address
10.1.1.1.
The SIP EA specifies the authentication parameters, which include the SIP registrar and SIP
proxy. "fhe connect and register timers are tuned to nondefault values.
The E'A uses the dial peer 10 to match all external destinations, points via SIP version 2 to the
ITSP SIP proxv. and uses I'CP as the transport protocol when signaling outbound calls.
Verifying SIP Gateways
This topic describes how to verity the operations of voice gateways in aSIP environment.
show sip-ua Command Overview
Command
show sip-ua service
show sip-ua status
show sip-ua regtoterstatus
show sip-ua timers
show sip-ua connections
show sip-ua calls
show sip-us statistics
DescripSon
Displays thestatusofthe SIPseivice
Displays the status ofthe SIP UA
Displays thestatusofE.164 numbers thata
SIP gateway hasreglstefadwith anexternal
primary SIP registrar
Displays SIP UAtimers
Displays active SIPUA connections
DisplaysactiveSIP UA calls
DisplaysSIP traffic statistics
The show commands that are listed in the table in the figure allow you to examine the status of
SIPcomponents andto troubleshool.
)2010 Cisco Systems. Inc.
VoIP Call Legs 2-103
SIP-UA General Verification
This subtopic explains how to verify general SIP UA status and settings.
2-104
S^P-UA General Verification
router# show eip-ua status
SIP User Agent status
SIP User Agent for UDP ; ENABLED
SIP User Agent for TCP : ENABLED
SIP User Agent (or TLS over TCP : ENABLED
SIP User Agent bind status [signal ing) : ENABLED 10.1.250.101
SIP User Agent Oind status(media): DISABLED
SIP early-media for 180 reapor.
ith SDP; ENABLED
SDP appil
nmespec line l"_ = ) required
Media supported, audio video image
Network types supported: IN
Address types supported: IP4 IP6
Transport types supported: RTP/AVP udptl
Ihe shovt sip-ua sen ice command displays the status of SIP call service on aSIP gatewav
Ihe sip-ua service isup when the VoIP service has not been shut down in the voice service
voip configuration mode. By default. VoIP service is enabled, and therefore SIP service is up.
The show sip-ua status command displavs the status for the SIP user agent. It shows which
transports are accepted for incoming calls. This output shows the default setting, which isto
accept UDP. TCP. and TCP TLS. Next, the interface binding information is displayed. In this
case, the signaling traffic is sourced from the address 10.1.250.101, and the media will be
sourced from the outgoing interface IP address. The command informs about the gatewav
support for SIP earlv media using 180 Ringing responses with SDP. It is enabled by default.
Ihe show sip-ua status command reports the required and supported SDP options'
Implementing Cisco Voice Communications andQoS(CVOICE) v80
SIP-UA Registration Status
This subtopic displays the F.I 64 numbers that arc registered on an external SIPregistrarserver.
SIP-UA Registration Status and Timers
router# show sip-ua register status
Line peer expires{sac) registered
4001 20001 596 no
4002 20002 596 no
51001 596 no
9998 2 596 no
router# show sip-ua timers
SIP DA Timer Values (millisecs)
trying 500, expires 180000, connect 500, disconnect 500
comet 500, prack 500, rellxx 500, notify 500
refer 500, register 500
fhe show sip-ua register status command displays the status of E.164 numbers that a SIP
gateway has registered with an external SIP registrar server. SIP gateways can register E. 164
numbers on behalf of analog telephone voice ports (FXS), IP phone virtual voice ports (EFXS).
and SCCP phones with an external SIP proxy or SIP registrar. The command show sip-ua
register status is only for outbound registration, so if there are no SCCP phones or FXS dial
peers to register, there is no output when the command is run. In this example, some endpoints
are attached to the SIP gateway, but they have not been registered with an external SIP
registrar.
The show sip-ua timers command displays the current settings for the SIP user-agent timers.
In this example, die command output shows the default values ofthe timers.
SIP-UA Call Information
This subtopic shows how to view the calls that are processed by a SIP UA.
SiP-UACail Information
router! alww aip-ua calls
SIP UAC CALL INFO
Number of SIP Ui[ Agent CllontfUAC] eel1b: 0
SIP UAS CALL INFO
Cell ;
SIP Cell in :D215P;;-7B5A11EJC-8005EJ>1A- 6ABP4ADS10 .10 ID.!
Stece of the ell . STATE ACTIVE CJ>
Calling Number J81S93I001
Celled HiBiêr : 1003
source IP Addrs. iSig 1. 10.10 10.1
De.tn SIP Req AddE.Port . 10.10 10.2:5060
Destn SIP P-esp Addi .Port . 10 10 . 10. 2 : 5 6 39e
[Mtinitloa Ku : 10.10.10.2
Wuuifcer of Hdi SlieuiB : 1
ItaMI of Active Stream. 1
Hedie Straui 1
suit ;i th* eram - stream active
Slr.j. Oil ID : 1
Striffi Type voice-only ID)
tiegotieted Codec . g72SrB (JO byteBi
Cwiee Peyloed Typ - IS
Necktie ted Dtui-teley : lDbind -voice
Kedi. Source IP Addr:Port: 10 10.10.1:190 50
Hodie Deet IP Addr:Port : 10.10.10.2:16522
fhe show sip-ua callscommand displavs active UAC and UAS calls and their parameters. The
output includes informationabout IPv6. Resource Reservation Protocol (RSVP), and media
forking (splitting the mediasession in multiple sessions) for each call on the deviceand for all
media streams associated with the calls. There can be any number of media streams associated
vsith a call, of which tvpically onlv one is active. A call can include up to three active media
streams if the call is media-forked.
2-106 Implemenling Cisco Voice Communications and QoS (CVOICE) v8 0
This subtopic explains how to debug SIP operations.
Command
debug ccsip
debug voip ccapi inout
Description
For general SlPdebugging; has
many detailed options, for example
for viewing direction-attribute settings
andportandnetwork address-
translation traces
Showseveryinteraction with the call
control API
debug voip ccapi protoheaders Displays messages sent between the
a originating and terminating gateways
The debug commands that are listed here are valuable when examining the status of SIP
components and troubleshooting:
debug ccsip command has various options as follows:
- debug ccsip all: This command enables all ccsip-type debugging. This debug
command is very active; you must use it sparingly in alive network,
- debug ccsip calls: This command displays all SIP call details as they are updated in
the SIP call control block. You must use this debug command to monitor call
records for suspicious clearing causes.
- debug ccsip errors: This command traces all errors that are encountered by the SIP
subsystem.
- - debug ccsip events: This command traces events, such as call setups, connections.
and disconnections. An events version ofadebug command is often the best place
tostart, because detailed debugs provide much useful information.
- debug ccsip info: This command enables tracing of general SIP security parameter
index (SPI) information, including verification that call redirection is disabled.
- debug ccsip media: This command enables tracing of SIP media streams.
debug ccsip messages: This command shows the headers of SIP messages that are
exchanged between a client and a server.
debug ccsip preauth: This command enables diagnostic reporting of authentication,
audiorization. and accounting (AAA) forSIP calls.
- debug ccsip states: This command displays the SIP states and state changes for
sessions within the SIP subsystem.
)2010 Cisco Systems. Inc
VoIPCall Legs 2-107
~ debug ccsip transport: This command enables tracing ofthe SIP transport handler
andtheTC Por UDP process.
debug voip ccapi inout: This command shows every interaction with the eall control
application programming interface (API) on both the telephone interface and on the VoIP
side B> monitoring the output, you can follow (he progress of acall from the inbound
interlace or \ olP peer to the outbound side ofthe call. This debug command is verv active-
you must use it sparingly in a live network.
debug voip ccapi protoheaders: This command displays messages that are sent between
the originating and terminating gateways. Ifno headers are being received bv the
terminating gateway. vcrify that the header-passing command is enabled on the
originating gateway.
2-108 Implemenling Cisco Voice Communicalions and OoS (CVOICE) v8 0
Examining the INVITE Message
This subtopic shows how to view the SIP INVITE message.
Examining the INViTE Message
router! debug ccsip messages
INVITE Hip:366 0210166. 34.2 45.2 3l!usei:"plione ; phone -contax tunknown
SIP/2.0
Via: SIP/2.0/UDP 166.34 .345.230:55820
From: 3660110- tsip:3660110*166.34.245.230>
To: <sip:3660210*16.34.245.23ljuser-phonoiphons-contaut-unknowns
Content-Type: application/adp
v-0
CiscoSyBtamsSIP-GH-oaaiAgant 4629 354 in ip4 55.1.1.42
s-SIP Call
e.IN IP4 55. 1. 1.42
t-0 0
m-audio 19978 RTP/AVP 0 100
c-IN IP4 10.1.1.42
artpmap:0 PCNU/SD0O
a-itpmapilOO Z USE/8000
The figure shows the output ofthe debug ccsip messages command. It shows the beginning of
a SIP INVITE message being sent from the endpoint with address 166.34.245.230 to the
endpoint with address 166.34.245.231. This example includes the description ofthe message
originator, the intended recipient, and, among other parameters, the content type, which is
application/sdp. The SDP description ofthe media capabilities is truncated in this output.
)20I0 Cisco Systems, Inc. VoIP Call Legs 2-109
Examining the 200 OK Message
I his subtopic explains how to examine the 200 OK message.
Examining the 200 OK IVlessi
route r# defcug ccsip messages
SIP/2 0 20 0 OK
Via; a IP/2 . 0/UDP 166.34 .245.230:55820
From : 3660110" <sip:3660110ei66.34,245.230s
To; < Bips3660210ei66.34.245.231fUBer.phon8;phone -
conteitunknowns;tag=27DHC6DS-13 57
Date : Hon, 08 Mar 1993 22:45:12 GMT
Call- ID: ABBJlE7AP-B23100E2-C-lCD274BCei72 18 . 192 . 194
Timeg tamp: 731427554
Serve r: Cisco VOIP Gateway/ IOS 12.x/ SIP enabled
Coat a ct : t sip: 36 0 210*166.34.245.231:5060 user^phone >
CSeq: 101 INVITE
Conte nt-Type: application/sdp
Conte nt-Length: 138
v=C
o=Cis coSystemsSIP-3W-UserAgent 1193 7927 IN IP4 166.34.245.231
s=SIP Call
t-0 0
C*IN P4 166.34.245.231
m=aud lO 20224 ETP/AVP 0
The ligure shows the output ofthe debug ccsip messages command. It shows a SIP 200 OK
message being sent in response to an earlier SIP INVITE7 message. The INVITI: message was
sent from 166.34,245.230 lo 166.34.245.23 1. and this address sel is retained in the 200 OK
message, with the addition ofthe Contact field that defines the originator of the 200 OK
message (166.34.245.231). The content ofthe 200 OK message includes, among other
parameters, the content tvpe. which is application/sdp. The second part ofthe output shows the
SDP description ofthe media. The media endpoint (the device that responds with the 200 OK
message) is 166.34.245.231. It will use UDP/RTP port 20224. The AVP is 0. which means the
call will use (i.711 mu-law.
Examining the BYE Message
This subtopic describes the BYE message.
Examining the BYE Message
router* debug ccsip messages
BYE sip;3660110166.34.245.230:5060iuser=phone SIP/2.0
Via: SIP/2.0/UDP 166.34.2*5.231:53600
Prom: <Hip:3660210166.34.245.231|User=plioneiphone-
context =un)cnown>; tag=27DBC6D8- 13 57
To: "3660110" eslp : 3660110*166 .34 .245. 230>
Date; (ton, 08 Mar 1993 22:45:14 GMT
Call-ID: ABBAE7AF-82 310 0E2-0-1CD274BC172.1B.192.19 4
User-Agent: Cisco VoIP Gateway/ IOS 12.x/ SIP enabled
Man-forwards: 6
Times tamp; 731612717
CSeq: 101 BYE
Content - Length: 0
The figure shows the output ofthe debug ccsip messages command. It shows the BYE
message that is sent when a call participant terminates the call.
Summary
Summary
SIP is a widely supported, still evolving signaling protocol.
SIP has three call setup models: direct gateway-to-gateway, connect over
proxy server and connect over redirect server
SIP address formats are FQDN, .164. and mixed.
SIP early media starts RTP flows before the call is answered
Basic SIP configuration may include UAsettings for communication over
a registrar or proxy server, and dial peers for direct gateway-to-gateway
calls
SIP ISDN support includes calling name display, blocking CLID when
privacy exists, and substituting the calling number for the display name,
if display name is unavailable
SIP secunty options relate to SIPS and SRST.
SIP gateway can be configured to use UDPn~CP/TLS transport, a specific
IP address, and modified timers
SIP operations can be verified using show and debug commands.
Implemenling Cisco Voice Communications and OoS (CVOICE) v8 0 2010 Cisco Systems, loc
Lesson 4
Explaining MGCP Signaling
Protocol
Overview
The Media Gateway Control Protocol (MGCP) enables the remote control and management of
voice and data communications devices at the edge of multiservice IP packet networks.
Because of its centralized architecture, MGCP overcomes the distributed configuration and
administrationproblems inherent in the use of protocols such as H.323. This lesson describes
how to configure MGCP on a gateway, and the features and functions ofthe MGCP
environment.
Objectives
Upon completing this lesson, you will be able to describe the characteristics of MGCPand
explain when to use it in a VoIP environment. This ability includes being able to meet these
objectives:
Describe MGCP, its components, and the advantages of MGCP as a voice gateway
protocol
Hxplain MGCPsignaling messages and the interactions between an MGCP call agent and
its associated gateways
Describe die process of codec negotiation in MGCP and explain how DTMF digits are
collected in MGCP
Configure an MGCP residential and trunking gateway
Describe how to configure the MGCP interlace binding and other parameters to conform to
the requirements ofthe call agent, trunks, or lines that are being used with the gateway
Describe major commands that are used to verify an MGCP gateway
MGCP Architecture
['his topic describes the MGCP, its architecture and operations.
Media Gateway Coi
Centralized device control with simple endpoints forbasicand
enhanced telephony services
An extension of Simple Gateway Control Protocol (SGCP) and
supports SGCP functionality in addition to several enhancements
a Allows remote control of various devices
* Stimulus protocol
Endpoints and gateways cannot function alone
Uses IETF SDP
Addressing by E.164 telephone number
* Defined in RFC 3435 and 2805
MGCP is an Internet irigineering I ask Force (ll.TF) -defined centralized device control
protocol. The MGCP protocol allows a central control component, or eall agent, to remotely
control various devices. The protocol is referred to as a stimulus protocol because the endpoints
and gatewavs cannot function alone. MGCP incorporates the IFTF Session Description
Protocol (SDP) to describe the tvpe of session to initiate.
MGCP is a plaintext protocol that uses a server-to-clienl relationship belween the call agent and
the gatewav to i'ullv control the gateway and its associated ports. The plaintext commands are
sent to gatewavs from the call agent using User Datagram Protocol (UDP) port 2427. Port 2727
is also used to send messages from the gateways to the call agent.
An MGCP gateway manages translation between audio signals and the packet network.
Gateways interact with a call agentalso called a media gateway controller (MGC)that
performs signal and eall processing on gateway calls. In the MGCP configurations that Cisco
[OS Software supports, a gateway can be a Cisco router, access server, or cable modem, and
the call agent is a server from a third-party vendor.
MGCP is an extension ofthe earlier version ofthe protocol Simple Gateway Control Protocol
(SGCP) and supports SGCP functionality in addition to several enhancements. Systems using
SGCP can easily migrate to MGCP, and MGCP commands are available to enable SGCP
capabilities. MGCP is similar to another standards-based protocol, Megaco.
2-114 Implementing Cisco Voice Communications and OoS (CVOICE) v8 0 2010 Cisco Systems Inc
MGCP Key Features
This subtopic explainsthe key features of MGCP.
MGCP Key Features
Alternative dial tone for VoIP environments (provider
competition)
Simplified configuration for dial peers
Simplified migration
Centralized dial plan configuration on the Cisco Unified
Centralized gateway configuration on the Cisco Unified
Simplified Cisco IOS configuration
Supports QSIGsupplementary services withCisco Unified
Uses UDP transport
There are several advantages to using MGCP-controlledgateways as voice gateways:
Alternative dial tone for VoIP environments: Deregulationin the telecommunications
industry gives competitive local exchange carriers(CLECs) opportunities to providetoll
bypassfrom the incumbent local exchangecarriers(ILHCs) with VoIP. MGCP enablesa
VoIPsystemto control call setupand teardown and CustomLocal AreaSubscriber
Services (CLASS) features for less sophisticated gateways.
Simplified configuration for dial peers: Whenyou use an MGCP call agent ina VoIP
environment, you do not need to configure static VoIP network dial peers. Residential
gateways need POTS dial peers, but their configuration is simplified. Trunking gateways
do not need any POTS dial peer configuration.
Migration paths: Systems usingearlierversionsofthe protocol can migrateeasily to
MGCP.
Centralized dial plan configured on the Cisco Unified Communications Manager: A
centralizeddial plan configurationon the Cisco UnifiedCommunications Manager enables
you to handle and manage the entire dial plan configurationon the Cisco Unified
Communications Manager cluster within a multisite network. This simplifies the
management and troubleshooting of a company telephone network.
Centralized gateway configuration on the Cisco Unified Communications Manager:
As in the case ofthe dial plan, centralized gateway configurations for all gateways arc
managed through one central configuration page, which simplifies the management and
troubleshootingof a company telephone network.
Caution Some network management tools donotwork correctly when performing theconfiguration
through Cisco Unified Communications Manager. In such cases, youmayneed to manually
configure the gateway for MGCP without using the configure network command.
Simple Cisco IOS gateway configuration: liecausethe gateway configuration is mostly
done on the Cisco UnifiedCommunications Manager, far fewer Cisco IOS router
commands are necessan to bring up the gateway than in any other gateway type.
Supports Q Signaling (QSIG) supplementary services with Cisco Unified
Communications Manager: With the support of QSIG supplementaryservices. MGCP is
a protocol that vou can use to interconnect a Cisco UnifiedCommunications Manager
environment with a traditional PBX.
Implementing Cisco Voice Communications and QoS (CVOICE! v8 0 2010 Cisco Systems, Inc.
MGCP Components
This subtopic describes the MGCP components.
IGCP Components
Cisco Voice
Gateways Can Agent (MGCP] Cisco Unified
Communications
Manager
I FXS= ForeignExchangeStatiwi
The distributed system is composed of a call agent (also called Media Gateway Controller), at
least one Media Gateway that performs the conversion of media signals between circuits and
packets, and at least one signaling gateway (SG) when connected to the PSTN.
MGCP defines a number of components and concepts. You must understand the relationships
between components and how the components use the concepts to implement a working MGCP
environment.
Here are the components that are used in an MGCP environment:
Call agent: A call agent exercises control over the operation of a gateway and its
associated endpoints by requesting that a gateway observe and report events. In response to
the events, the call agent instructs the endpoint what signal, if any, the endpoint should
send to the attached telephone equipment. This requires a call agent to recognize each
endpoint type that it supports and the signaling characteristics of each physical and logical
interface that is attached to a gateway. The call agent can audit the current state of
endpoints on a gateway. A call agent uses its directory of endpoints and the relationship
that each endpoint has with the dial plan to determine appropriate call routing. Call agents
initiate all VoIP call legs.
Gateways: Gateways manage the translation of audio between the SCN and the packet
network. The gateway uses MGCP to report events (such as off-hook or dialed digits) to the
call agent. The two main types of gateways are residential and trunking gateways. The
main types of MGCP gateways are as follows:
Residential gateway: A residential gateway provides an interface between analog
(RJ-11) calls from a telephone and the VoIP network. The interfaces on a residential
gateway may terminate a plain old telephone service (POTS) connection to a phone
or a PBX.
Trunking gateway: A trunking gateway provides an interface between PSTNtrunks
and a VoIP network. A trunk can be a digital service level 0 (DSO). Tl. or Rl line.
Kndpoints: Endpoints represent the point of interconnectionbetween the packet network
and the traditional telephone network.
The figure slums an MGCP env ironment withall threecomponents. Ciscovoice gateways act
as MGCPgatewavs. and Cisco UnifiedCommunications Manager acts as the MGCPcall agent.
2-118 Implemenling Cisco Voice Communications and QoS (CVOICE) v8.0 2010 Cisco Systems, Inc
MGCP Gateways
Thissubtopic describes theroleof MGCP gateways.
MGCP Gateways
Call processing is done by a call agent suchas Cisco Unified
Communications Manager.
MGCPuses endpoints and connections to construct a call.
- Endpoints:
Sources of data or destinations for data
- Physical or logical locations in a device
- Connections:
Point-to-point
Multipoint
InMGCP, the call agent plays acentra! role by controlling the setting up and tearing down of
connections between theendpoints ina VoIP network andendpoints in the PSTN, while
managing all dial-plan configuration elements, The calls are routed via route patterns onthe
call agent, not by dial peers onthe gateway. The gateway voice ports must beconfigured for
proper signaling. There are no dial peers for MGCP except when arouter isusing Cisco
Unified Survivable Remote Site Telephony (Cisco Unified SRST)for fallback.
MGCP uses endpoints and connections toconstruct a call. Endpoints aresources ofdata or
destinations for data and canbe physicalor logicallocations ina device.
MGCP gateway connections can bepoint-to-point ormultipoint. Apoint-to-point connection is
anassociation between twoendpoints with thepurpose of transmitting databetween these
endpoints. Data transfer between these endpoints can take place after this association is
established for both endpoints. Amultipoint connection isestablished byconnecting the
endpoint toa multipoint session. Connections can beestablished overseveral types of bearer
networks:
Transmission of audio packets using Real-Time Transport Protocol (RTP)/UDP overan IP
network.
Transmission of packets over aninternal connection. This method is used, inparticular, for
"hairpin" connections thatareconnections that terminate ina gateway butareimmediately
rerouted over the telephone network.
MGCP Endpoints
This subtopic explains the MGCP endpoints and their identifiers.
SVSU1/DSM@gw1 .domain.com
AALN/S2/SUV1@gw1.domain.com
Subunrt 1
v..MS2/tfU1/1@flwS.domain com
Port 1
Porl 1
iVSU1/r)S'-1@aw1 dom*fi com
When interacting with a gateway, the call agent directs itscommands tothe gateway for
managing anendpoint or a group of endpoints. An endpoint identifier, as itsname suggests,
identifies endpoints.
Hndpoint identifiers consisl of two parts: a local name ofthe endpoint inthecontext ofthe
gatewav andthedomain name ofthe gateway itself. Thetwopartsareseparated by an"at" sign
(" a"). If the local part represents a hierarchy, thesubparts ofthe hierarchy arcseparated bya
slash (/). Inthe figure, the local IDmav berepresentative of a particular "gateway/circuit #."
and the "circuit *"ma\ in turn be representative of a "circuit ID/channel #".
2-120 Implementing Cisco Voice Communicalions and QoS (CVOICE] v8 0 >2010Cisco Systems, Inc
MGCP Package Types
This subtopic explains the MGCP package types.
MGCP Package Types
Groups of event and signal definitions:
- Compatibility
Modularity
* Enabledwith the mgcp package-capability command:
- Trunk
- Line
- Dual-tone multifrequency (DTMF)
- Generic media
- Real-Time Transport Protocol (RTP)
- Announcement server
- Script
Creating a call connection involves aseries ofsignals and events that describe the connection
process. Each event causes signal messages tobe sent tothe call agent, and associated
commands aresent back. The signals and events thatcompose theconnection process might
include indicators suchas theofT-hook eventthattriggers a dial tonesignal. These events and
signals are specific tothe type ofendpoint that isinvolved in the call. MGCP groups these
events and signals into packages.
Atrunk package, for example, isagroup ofevents and signals relevant toa trunking gateway;
an announcement package is a group of events andsignals relevant to anannouncement server.
These packages arc enabled by using the mgcp package-capability command. MGCP supports
thefollowing seven package types using theprovided command example:
Trunk: mgcp package-capability trunk-package
Line: mgcp package-capability line-package
Dual tone multifrequency (DTMF): mgcp package-capability dtmf-package
Generic media: mgcp package-capability gm-packagc
RTP: mgcp package-capability rtp-packagc
Announcement server: mgcp package-capability as-package
Script: mgcp package-capability script-package
Thetrunk package andline package aresupported bydefault on certain types of gateways.
Although configuring a gateway with additional endpoint package information isoptional, you
maywant tospecify packages foryourendpoints to addinformation, or youmay wantto
override the defaults on some ofthe packages.
MGCP Call Flows
This topic describes the process ofsetting up and tearing down calls, and explains the messages
that are involved in this process.
Audit End point (AUEP)
AuditConnection (AUOQ
Endpoin Configuration (EPCF)
CreateConnection (CRCX)
ModifyCoroectton (MDCX|
DeleteConnection (DLCX)
NolocationRe quest (RQNT)
Notify (NTFY)
RestartinProgress (RSP)
Call agent requests thostalus of an endpoint
Call agenl requests the status o(a connection
Call agent instructsihe gateway about the codingcharacteristics
expected by the "line-side"'ol the endpoint
Call agent instructs trie gateway to establish a correction wrth an
endpoml
Call agent instructsthe gateway to update its connection parameters
for a previously established connection
Gateway or call agent reports that it no longer has the resources to
sustain the call and intorms the recipient to delete connection
Call agent instructs the gateway to watch for events on an endpolnl
and the action to take when they occur
Gateway informs the call agent of an event for which notification was
requested
Gateway notifies the call agent that the gateway and rls endpoints are
removed from service or are being placed hack in service
MGCP packets are unlike what is found in manv other protocols. Usually wrapped in UDP port
2427. the MGCP datagrams are formatted withwhitespace, similarto what you wouldexpect
to find in TCPprotocols. An MGCP packet is either a command or a response.
Acall agent usescontrol messages to direct its gateways and their operational behavior.
Gatewav s use the control messages inresponding to requests from a call agent and notifying
the call agent of events and abnormal behavior.
Ihcreare multiple MGCP messages, twoof which arc used bv a call agent toquery Ihe state of
a Media Gatewav:
Auditl'ndpoint (Al'KP): fhis message requests the status of an endpoint. The call agent
issues the command. The following endpoint info can be audited with this command:
KequestedHvcnts. DigitMap. SignaiRequests. Rcquestldenlilier, QuarantineIIandling.
Notiliedfintity. Conneetionldentiliers. DetectEvenls. Observed Events, FvcntStates,
Bearerlnformation. KestartMethod. Rcstartl>ela>. ReasonCode, Packagel.isl.
MaxMGCPDatagram. and Capabilities. The response will in turn include infonnation about
each ofthe items for which auditing info was requested.
Autoconnection (Al'CX): This message requests the status of a connection. 1he call
agent issues the command.
Implementing Cisco Voice Communications and QoS (CVOICE) vS.O 2010 Cisco Systems, Inc
One message is used bya call agent tomanage a media gateway:
tndpointConfiguration (El'CF): The EndpointConfiguration command can be used to
specifv the encoding ofthe signals that will be received by the endpoint. For example, in
certain international telephony configurations, some calls will carry mu-law encoded audio
signals, while others will use A-law. The Call Agent can use the EndpointConfiguration
command to pass this information tothegateway. Theconfiguration may vary ona call-by-
call basis, but can also be used in the absence of any connection.
Three messages are used bya call agent to manage an RTP connection on a Media Gateway. (A
Media Gateway can also send a DUCX when it needs todelete a connection for itsself-
management):
CreateConnection(CRCX): fhis message instructs thegateway to establish a connection
withanendpoint. Thecall agent issues thecommand. Aconnection is defined by its
endpoints. Theinput parameters inCreateConnection provide the data necessary to build a
gateway "view" of a connection. Theparameters include thecodec, packetization period.
foSmarking, usage of echo cancellation, silence suppression, gain control, RTP security,
and resource reservation.
DeleteConnection(DLCX): This messageinformsthe recipientto deletea connection.
fhe call agent or the gateway can issuethe command. Thegateway or the call agent issues
the command to advise that it no longer has the resources to sustain the call. As a side
effect, the call agent collectsstatisticson theexecution ofthe connection. The statistics
include numberof packets sent, received, and lost, interarrival jitter, and average
transmission delay.
ModifyConnection(MDCX): This messageinstructs thegatewayto updateits connection
parameters for a previously established connection. The call agent issuesthe command.
The parameters usedare the sameas in the CreateConnection command, withthe addition
of a Conncctionld that identifies the connection within the endpoint.
One message is usedby a call agent to request notification of eventson the MediaGatewayand
to request a Media Gateway to apply signals:
NotificationRequest (RQNT): This messageinstructs the gateway to watch for eventson
an endpoint andthe actionto take whenthey occur. For example, a notification may be
requested for when a gateway detects that an endpoint is receiving tones associated with
fax communication. The entity receiving this notification may then decide to specify use of
a ditTerent type of encoding method in the connections bound to this endpoint and instruct
thegatewayaccordingly witha ModifyConnection command.
One message is usedby a MediaGateway to indicate to the call agent that it has detectedan
event for which notification was requested by the call agent (via the RQNT message):
Notify(NTFY): This message informs the call agent of an event for whichnotification was
requested, 'fhis message carries a list of events that the gateway detected and accumulated.
Asingle notification may report a list of events that will be reported in the order in which
they were detected (FIFO).
One verb isused bv a Media Gatewav to indicate tothe call agent thai it isinthe process of
restarting:
RestartlnProgress (RSIP): Thismessage notifies thecall agent that thegateway andits
endpoints are removed from service or are being placed back inservice, fhe gateway
issues the message. Themessage carries an EndPointld to identify theendpoint(s) that are
put in-serv iceor out-of-scrv ice. The RestartMethod parameter specifies the type of restart.
The following values have been defined:
"Graceful" restart method indicates that the specified endpoints will be taken out-of-
scrvice after the specified delav.
"forced" restart method indicates that the specified endpoints arc taken abruptly out-
of-servicc. The established connections, if any. are lost.
"Restart" method indicates that servicewill be restored on theendpoints after the
specified "restart delav."that is. the endpoints will be in-service. The endpoints are in
their clean default state and there are no connections that arc currently established on
the endpoints.
"Disconnected" method indicates that ihe endpoint has become disconnected and is
now tn ing lo establish connectivitv. The "restart delay" specifies the number of
seconds the endpoint has been disconnected. F.slablished connections are not
affected.
"Cancel-graceful"method indicates that a gateway is canceling a previously issued
"graceful" restart command. The endpoints are still in-service.
2-124 Implementing Cisco Voice Communications and QoS (CVOICE) v8 0 2010 Cisco Systems Inc
Residential Gateway to Residential Gateway
This subtopic describes the process ofsetting up and tearing down calls when two residential
gateways communicate over thecall agent.
MGCP Call Flows
Residential Gateway to Residential Gateway
The figure illustrates a dialogbetween a call agent and two residential gateways:
Step 1 The call agent sendsan RQNTto each gateway. Because they are residential
gateways, the request instructs the gateways to wait for an off-hooktransition
(event). When the off-hook transition event occurs, the call agent instructs the
gateways tosupply a dial tone(signal), 'fhe call agent asksthegateway tomonitor
for odier events as well. By providing a digit map in the request, the call agent can
have the gateway collect digits before il notifies the call agent.
Step 2 The gateways respond to the request. At this point, the gateways and the call agent
wait for a triggering event.
Step3 Auser on gateway Agoes off'hook. As instructed by the call agent in its earlier
request the gateway provides a dial tone. Because thegateway is provided witha
digit map. it begins to collect digits (as they are dialed) until either a match is made
or no match is possible. For the remainder of this example, assume that the digits
match a digit map entry.
Step 4 Gateway A sends an NTFY to the call agent to advise the call agent that a requested
event was observed, 'fhe NTFY identities the endpoint, the event, and in this case,
the dialed digits.
Step 5 After confirming that a call is possible based on the dialed digits, the call agent
instructs gateway A to create a connection (by sending CRCX) with its endpoint.
Step 6 The gateway responds with a session description if it is able to accommodate the
connection. The session description identifies at least the IP address and UDP port
for use in a subsequent RTP session. The gateway does not have a session
description for the remote side ofthe call, and the connection enters a wait state.
Step 7 The call agent prepares and sends a connection request togateway B. Inthe request,
the call agent prov ides the session description that isobtained from gateway A. The
connection request is targeted lo a singleendpoint (if onlyone endpoint is capable of
processing the call) or to anv one of a set of endpoints. The call agent alsoembedsa
notification request that instructs the gateway about the signalsand eventsthat it
should now considerrelevant. In this example, wherethe gateway is residential, the
signal requests ringing and the event is an off-hook transition.
Note The interaction between gateway Band its attached user has been simplified.
Step8 Gatewav Bresponds to the request with its session description. Noticethat gateway
B has both session descriptions and recognizes how lo establish its RTP sessions.
Step 9 fhe call agent relays the session description to gateway A in an MDCX. This
request may contain an encapsulated NTFY request that describes the relevant
signals and events at this stageofthe call setup. Now gateway Aand gateway B
have the required session descriptions to establish the RIP sessions over which the
audio travels.
Step 10 At the conclusion ofthe call, one ofthe endpoints recognizes an on-hook transition.
In the example, the user on gatewav A hangs up. Becausethe call agent requested
the gateways to notify the call agent in such an event, gateway A notifies the call
agent.
Step 11 The call agent sends a Dl.CX requesl to each gateway.
Step 12 The gatewavs delete the connections and respond.
2-126 Implementing Cisco Voice Communications and QoS (CVOICE) vS.O 2010 Cisco Systems, Inc.
Trunking Gateway to Trunking Gateway
This subtopic describes the process ofsetting up and tearing down calls when two trunking
gateways communicate overthecall agent.
MGCP Call Flows (Cont)
Trunking Gateway to Trunking Gateway
Gateway^
CRCX (SDP) i
'CRCX Response (SDP)!
Q931 Setup
i '
! Q931 Alerting ;
] !
' !
Q931 Connect ;
"fhe figure illustrates a dialog between a call agent and twotrunking gateways:
Step1 AQ931 Setup message comes in tothecall agent on anISDN trunk. Thisaction
triggers the call agent toanalyze the Setup message and decide how tomanage it by
consulting its local configuration. Thelocal configuration tellsthecall agent toset
upa call between two specific endpoints onGateways Aand B.
Step 2 The call agent sends two CRCX messages thataresimilar to the call flow illustrated
for residential-to-residential scenario.
Step3 Both gateways respond with a session description thatcontains the IPaddress and
UDPport for use inthe subsequent RTPsession.
Step 4 When the destination endpoint seizes theline, the destination central office (CO)
switch sends the Q.931 Alerting message to the eall agent.
Step 5 The call agent sendsan MDCX to gateway A. The ConneetionMode is nowset to
recvonly. Thisis because, at thispoint, only theringing loneneeds to flow back to
the originating side. Theoriginating side cannot sendaudioyet.
Step 6 Afterthe terminating phoneis pickedup, the destination COsendsthe Q931
Connect message to the call agent.
Step7 The call agent sends an MDCX logateway A, setting themode tosendrccv. This
step completes the setup of gateways for two-way audio.
MGCP Special Considerations
This topic describes the ISDN PRI Backhaul feature, the codec negotiation, and the digit
collection process in MGCP.
PRI Backhaul
D-channel call-setup signals
need to be earned in their
raw form back to the Cisco
Unified Communications
Manager to be processed
Gateway terminates data
link layer and passes the
rest of signals (0 931 and
above) to Cisco Unified
via TCP port 2428.
D-channel will be down
unless it can communicate
with Cisco Unified
Cisco Unified
Communications
Manager
A PRI backhaul is an interna] interface between Cisco Unified Communications Manager and
CiscoMGCP galewav s (that is. a separate channel lor backhauling signaling infonnation). It
forwards l.aver 3 PRI (0-931) backhaulcdover a TCP connection. Layer 3 informationis
forwarded independent of the native protocol that is used on the PSTN time-division
multiplexing (TDM) interface.
A PRI is distinguished from other interfaces by ihe fact that data received from the PSTN on
the D-channel must be carried in its raw form back to fhe Cisco Unified Communications
Manager to be processed, 'fhe gateway does not process or change this signaling data, it simply
passes it onto the call agent through I'CPport 2428. The gateway is still responsible for the
terminationofthe l.aver 2 data. That means that all the Q.921 data link layer connection
protocols are terminatedon the gatewav. bul evcrv'lhing above that (Q.931 network layer data
and beyond) is passed ontothe call agent. This also means that the gateway does not bringup
the D-channel unless it can communicate with Cisco UnifiedCommunications Manager to
backhaul the Q.931 messages, contained in the D-channcl. The figure illustrates these
relationships.
2-128 Implementing Cisco Voice Communications and QoS (CVOICE) v8 0 )2010 Cisco Systems. Inc
Codec Negotiation (Residential Gateway-to-Residential
Gateway Example)
This subtopic explains the codec negotiation in MGCP. This example shows how residential
gateways communicate over a call agent.
Codec Negotiation (Residential
Gateway-to-Residential Gateway)
Cffl Hook and
Dialed 5551234
MGCP uses SOP lo describe
and negotiate media
parwnetera: RTPjRTCP ports
and codecs.
Codec proposals are sen!
OGW->CA->TGW.
Codec corflkmaUon is sen
TGW->CA-OGW.
On Hook
DLCX
CRCX (SOP.
npcfl Pponse CSPP); Encapsulated
-DICX_
Ringing,
Then Answer
MGCP supports
early media. Ifearly
media is negotiated,
media channel is
established before
the call is accepted.
This subtopic explains the codec negotiation in MGCP between residential gateways. The
figure showstwo residential gateways communicating over a call agent, but the sameprocess is
involved with other gateway combinations. Like in Session Initiation Protocol (SIP), MGCP
gateways use SDPtoexchange mediacapabilities. The figure illustrates the SDPmessages that
carry the mediacodec information, and RTP/UDP port numbers. The codecproposalsand port
numbers ofthe originating gateway are sent in the CRCX response lo the first CRCXmessage,
and then forwarded to the terminatinggateway in the subsequent CRCX message. The
terminating gateway encapsulates the selectedcodecand its RTPport numbers inthe CRCX
Response lo the call agent, and this information is then forwarded to the originating gateway in
the MDCX message.
Like 11323 and SIP, MGCP supports early media. If early media is successfully negotiated, the
gateways start streaming the media before the call is answered.
Digit Collection
This subtopic explains ihe digit collection process in MGCP
Digit Collection
Gait-
AgBftt:
=' Notilication Requesl (RQNT)
NTFY (Off-hookl
Notification Requesl (RQNT)

Off-hook
and dialed
1234
NTFY (It Call agent keeps
requesting digits
until it finds a
dial plan match
Nolificalion Request (RQNT)
NTFY (21
NTFY (31
Notification Request (RQNT)
NTFY Ml
Create Connection (CRCX)
CRCX ResDonse
fhis figure illustrates a call agent controlling a residential gateway. Alter the gateway registers
the endpoint on the call agent, the call agent sends a notification request (RQNT) to the
gatewav. This message instructs the gateway lo watch for events on an endpoint. When the user
takes the lelephone off-hook, the gateway reports this event using the NTFY message. With the
endpoint taken off-hook, the call agent consults its dial plan, and if an entry for an empty dial
string exists, the call is forwarded to the defined destination. In this example, the call uses a
four-digit-based dial-plan and the call agent sends another RQNT, telling the gateway to watch
for further events on the endpoint. The user dials the first digit (I), which is reported in a NTFY
message to the call agent. The call agent keeps checking its dial plan and requesting further
events until a complete routable number is collected, "fhe call agent then instructs the gateway
to create the connection using the previously explained procedure.
2-130 Implementing Cisco Voice Communications and QoS (CVOICE) v8 0 )2010 Cisco Systems. Inc
r-5
Configuring MGCP Gateways
This topic describes how toconfigure MGCP gateways.
MGCP Gateway Configuration Overview
Tvao scenarios:
Residential gateway
Dial peer configuration
Trunkinggateway
- Controller configuration
The configuration ofMGCP gateways varies based onthe role that they play inthe network:
Residential gateways are configured using dial peers.
Trunking gateways areconfigured usingcontroller settings.
)2010 Cisco Systems. Inc
MGCP Commands
fhis subtopic describes the MGCP commands that are used for basic MGCP configuration on
the voice aateuavs.
MGCP Commands
router(config)#
mgcp [port]
Starts the MGCPdaemon on specified UDP port (default 2427)
Allocates resources to MGCP processes
router(config)#
corn-manager mgcp
- Enables MGCP communications with Unified CM*
Enables redundancy when a backup Unified CM* is available
router(config)#
mgcp call-agent {host-name I ip-address} [port] [service -
type type [version protocol-version]]
Defines the address and protocol of the call agent
- Service type set to mgcp
' Unrfiea CM = Ctsco Unrfied Communications Manager
Threecommands arc needed to configure basic MGCP gateway functionality.
flic mgcp command allocates resources to the MGCPprocess and starts the MGCPdaemon,
"fhe optional port valuespecifies the UDPport that the MGCP daemon is listening on. The
default is UDP port 2427.
The ccm-manager mgcp command enables communications with Cisco Unified
Communications Manager and activates support for redundancy whenmultiple Cisco Unified
Communications Managers are available.
The mgcp call-agent command configures the address and UDP port ofthe call agent. The
optional service-type parameters can define one of these protocols: MGCP. Network-Based
Call Signaling (NCS). SGCP. or frunking Gateway Control Protocol (TGCP). The version of
MGCP can take these values: 0.1 (MGCP Internet Draft). 1.0 (MGCP RFC 2705 version 1,0).
RFC 3435-1.0- Version 1.0 of MGCP(RKC 3435 version 1.0). The default values are as
follows:
Port: UDP port 2427
Service type and version: MGCP p 0.1
Customizing MGCP Gateways
Thistopic describes how tocustomize MGCP gateways.
Configuration Download and Source Address
router(config)#
Iccm-manager config server {ip-address | name}]
Specifies the TFTP server address
Downloads Cisco Unified Communications Manager XML
configuration files
router (conf ig) It
Imgcp bind {control | madia} aourca-interfaea int-id
Sets the source address for signaling and media packets to the
IP address of a specific interface
Ifnot configured, the IP layer selects the best address
The customization of MGCP settingson a voice gateway may includea wide rangeof settings.
Tosimplify the provisioning process, theCisco Unified Communications Manager offers a
TFTPservicethat the gateways can use to download the appropriate parameters.
TTie ccm-managerconfigserver command specifies theTFTP address from which to
download thesettings andstartsthedownload process. Thiscommand is frequently used to
automate the gateway customizationprocess.
The mgcp bind command enablesthe interface bindingfeature that allows the MGCP process
to use a specific interface IPaddress. Either signaling, or media, or bothsignaling andmedia
can be sourced from the address ofthe defined interface. If the mgcp bind command is not
configured, the IPlayerselects theaddress, based on theoutgoing interface intherouting
decision.
Package Configuration
This subtopic explains how to configure theuse of MGCP packages.
Package Configuration
router(config)#
jmgcp default-package package
Configure the default package capability type
Default for residential gateways: line-package
Defaultfor trunking gateways trunk-package
router Iconfig) #
mgcp package-capability package
Configure additional package capability
In addition to the default package
Package should be supported by the specific call agent
1he mgcp default-package command sets thedefault package capability that describes the
basicset of eventsand signalsthat the galewav uses, flic default for residential gateways is
line-package.
1he mgcp package-capabilitycommand specifies the MGCP package capability type for a
media gatewav, ! his command is useful for defining extension packages in addition to the
default package.
2-134 Implementing Cisco Voice Communications and QoS (CVOICE) vB.O 2010 Cisco Systems. Inc
Selected Package Types
This subtopic lists the package types.
Package Types
line-package Package or residential lines; default for residential gateways
trunk-package Events and signals for trunk lines; default for trunking gateways
as-package Announcement server package
script-package Events and signals for script loading
srtp-package Secure RTP (SRTP) package; the default is disabled
dt-package Events and signals for immediate-start, DTMF and dial-pulse trunks
dlmf-package Events and signals for DTMFrelay
fxr-package Events and signals for fax transmissions
gm-package Events and signals for several types of endpoints, such as trunking
gateways, access gateways, or residential gateways
md-package Provides support for Feature Group D (FGD) Exchange Access North
American (EANA) protocol signaling
ms-package Events and signals for wink-start and immediate-start DIDand Direct
Outward Dialing (DQD),basic Rl, and FGDTerminatingProtocol
fhis table lists some ofthe available package types and their descriptions.
Residential Gateway Example
This subtopic presents the residential gatewav customizationexample.
Cisco Unrfied
Communicalions
Manager
10111
IP
com manager mgcp
mgcp
mgcp call-agent 10.1.1.1
mgcp bind control source-interface FastethernetO/0
mgcp bind media source-interface FastethernetO/0
dial-peer voice 1 pote
service mgcpapp
port 1/0/0
mgcp package-capability dtmf-package
mgcp package-capability gin-package
mgcp package-capability line-package
mgcp default-package line-package
fhe first three commands enable the galewav lo inleropcrale with Cisco Unified
Communications Manager, start the MGCP daemon, and define the Cisco Unified
Communications Manager address.
"fhe signaling and media tralfic is sourced from the IP address ofthe fastethernetOVO interface.
Further, the POTS dial peer configuration:
Specifies the MGCP application to run on the voice port.
Defines the voice port to bind with MGCP.
When the parameters ofthe MGCP gateway are configured, the active voice ports (endpoints)
are associated with the MGCP. The application mgcpapp command, issued in the dial peer
configuration mode, binds the appropriate voice port to MGCP. The dial peers do not have a
destination pattern. A destination pattern is not used because the relationship between the
dialed number and the port is maintained by the call agent.
In addition to the basic settings, the package capabilities are extended by three additional
packages: dtmf-package. gm-package. and line-package.
2-136 Implementing Cisco Voice Communications and QoS (CVOICE) v8 i >2010 Cisco Systems, Inc
Trunking Gateway Example
Thissubtopic provides thecustomization example ofthe trunking gateway.
Trunking Gateway Customization Example
Cisco Unrfied
Communications
Manager
101 1 1
network-clock-aelect 1 el 0/1/0
ecm-manager mgcp
mgcp
mgcp call-agent 10.1.1.1
mgcp bind control source-interface FastetharnetO/0
mgcp bind media eource-interface FaotethernetO/0
controller tl 0/1/0
da0-group 1 timeslots 1-24 service mgcp
I
mgcp default-package trunk-package
mgcp package-capability dt-package
The figure shows a trunk gateway configuration and customization example. Insteadof using
the application mgcpapp command in a dial peer, a trunk endpoint identifies its association
with MGCP using the service mgcp parameter in the dsO-group controller subcommand. As
always in MGCP, the call agent maintains the relationshipbetween the endpoint (in this case, a
digital trunk) and its address.
In addition to the basic settings, the signaling and media traffic is sourced fromthe IP address
ofthe FastethemetO/0 interface. The package capabilities are extended by one additional
package: dt-package.
Verifying MGCP Gateways
his topic describes howto veritv the MGCPgateway operations.
show mgcp Command:
Command
show mgcp
show ccm-manager
show mgcpendpoint
show mgcp statistics
Displays MGCP status and parameters
Displays the registration state with the
primary, secondary, and tertiary call agent
Displays the MGCP endpoint names, related
voice ports, and administrative status
Shows MGCP packet statistics
The common commands that are used for verify ing MGCP gateways are the following:
show mgcp: Displavs MGCP status and parameters.
show ccm-manager: Shows the registration state with the primary, secondary, and tertiary
call agent.
show mgcp endpoint: Provides the MGCP endpoint names, related voice ports, and
administrative status.
show mgcp statistics: Displavs MGCP packet statistics.
2-138 Implementing Cisco Voice Communications and QoS (CVOICE) v8 0 i2010 Cisco Systems, Inc.
show mgcp Command
This subtopic presents the sample output oftheshow mgcp command.
show mgcp Command
router# show mgcp
MGCP Admin state active, Oper State active - Cause Code HONE
MGCP call-agent: 10.1.1.101 4000 Initial protocol service is
MGCP 0.1
MGCP media (RTP) dscp; ef, MGCP signaling deep: af31
MGCP default package: trunk-package
MGCP supported packages! gin-package dtmf-package trunk-package
line-package hs-pac)cage atm-package ms-package dt-package mo-
package res-package mt-package fxr-package md-package
mgcp Digit Map matching order; shortest match
SGCP Digit Hap matching orden always left-to-right
MGCP control bind :DISABLED
MGCP media bind :DISABLED
MGCP Dpspeed payload type tor GTllulawj 0, G711alawt 8
MGCP Dynamic payload type for G.726-16K codec
MGCP Dynamic payload type for G.726-24K codec
MGCP Dynamic payload type for G.Clear codec
This figure provides the sample output ofthe showmgcp command. The command displays
the administrative stateofthe gateway, theaddress ofthe call agent, information about the
loaded packages, thestatusofthe interface binding feature, anda list of supported codecs.
show ccm-manager Command
This subtopic presents the sample output ofthe show ccm-manager command.
show ccm-manaqer Co
router# show ccm-manager
MGCP Domain Name: cisco-voice-01
Priority Status
Primary
First Backup
Second Backup
Registered
None
None
Current active Call Manager: 10.1.1.1
Backhaul/Redundant link port: 2428
Failover Interval: 30 seconds
Keepalive Interval: IE seconds
Last keepalive sent: 5wld (elapsed time: 00:00:04)
Last MGCP traffic time: Swld (elapsed time: 00:00:04)
Last Cailover time: None
Switchback mode: Graceful
This figure provides the sample output of the show ccm-managcr command. The command
displavs the domain name, which will be used when constructingthe names for endpoints
registered on the call agent. Further, the command provides the registrationstatus and IP
addresses ofthe first three call agents, the currently active call agent, the port that is used for
the Backhaul feature, and some timing characteristics.
show mgcp endpoint Command
This subtopic presents the sample output ofthe show mgcp endpoint command.
show mgcp endpo nt Command
routerll show mgcp endpoint
Interface Tl 0/1/0
ENDPOINT-NAME V-PORT SIG-TYPE ADMIN
SO/SDl/dSl-0/lHQ-l 0/1/0:1 none up
S 0/SOI/dSi-0/2HQ-1 0/1/0:1 none up
S0/SOl/d8l-0/39HQ-l 0/1/0:1 none up
S0/SOl/dSl-0/48HQ-l 0/1/0:1 none up
S0/sni/del-0/5HQ-i 0/1/0:1 none up
SO/SUl/del-0/6HQ,-l 0/1/0:1 none up
SO/SUl/dsl-0/79HQ-l 0/1/0:1 none up
S0/SDl/dsl-0/BHQ-l 0/1/0:1 none up
S0/SOl/dSl-0/9HQ-l 0/1/0:1 none up
S0/SDl/dSl-O/10HQ-l 0/1/0i1 none up
This figure provides the sample output ofthe show mgcpendpoint command. Thecommand
displays a list ofthe voiceportsthat are configured for MGCP.
12010 Cisco Syslems. Inc. VoIP Call Legs 2-141
show mgcp statistics Command
This subtopic presents the sample output ofthe showmgcp statistics command.
show mgcp statistics G
router# show mgcp statistics
TOP pkts rx 8, tx 9
Unrecognized rx pkts 0, MGCP message parsing errors 0
Duplicate MGCP ack tx 0, Invalid versions count 0
CreateConn rx 4, successful 0, failed 0
DeleteConn rx 2, successful 2, failed 0
ModifyConn rx 4, successful 4, failed
DeleteConn tx 0, successful 0, failed a
NotifyRequest rx 0, successful 4, failed 0
AuditConnection rx 0. successful 0, failed 0
AuditEndpoint rx 0, successful a, failed 0
RestartlnProgress tx 1, successful 1. failed 0
Notify tx 0, successful 0, failed 0
ACK tx a, NACK tx 0
ACK rx 0, NACK rx 0
IP address based Call Agents statistics:
IP address 10.1.1.1, Total msg rx 8, successful 8, failed 0
Ihis figure provides the sample output oflhe show mgcp statistics command, fhe command
displavs a count ofthe successful and unsuccessful control commands. A high number of
unsuccessful (failed) messages indicates a communication problem with the call agent.
2-142 Implementing Cisco Voice Communicalions and QoS (CVOICE) v8 i '2010 Cisco Systems. Inc
Summary
Summary
* MGCP is a server/client signaling protocol that allows the call
agent to control the MGCPgateway.
* Call agent sends instructions tothe MGCP gateway andthe
gateway responds.
* MGCPcodec selection is dictated by the call agent, and defined
on the Cisco Unified Communications Manager using the
regions approach.
Two maintypes of MGCP gateways are residential and trunking.
* MGCP gatewayloads packages to supportevents and signals.
* MGCPstatus, registration, endpoints, and statistics can be
viewed using appropriate show commands.
2-144 Implementing CiscoVoice Communications and QoS (CVOICE) v80 2010CiscoSystems, Inc
Lesson 5
Describing Requirements for
VoIP Call Legs
Overview
The inherent characteristics of a converged voice and data IP network create challenges for
network engineers andadministrators indelivering voice traffic, 'fhis lesson describes the
challenges of integrating a voice anddatanetwork andofTers solutions fordesigning a VoIP
network for optimal voice quality, fax, and modemtransmission.
Objectives
Upon completing this lesson, youwill beabletodescribe special requirements for VoIP call
legs including the need for qualityof service(QoS), faxand modem relay, and dual tone
multifrequency (DTMF) support. This abilityincludes beingable to meet these objectives:
Describe the factors present in IP networks that affect audio clarity
Describe the QoSrequirement of VoIPand QoS features as they relate to a VoIPnetwork
Describe the challenge of transporting fax and modem calls over IP networks
F.xplain howfax/modem pass-through, relay, and store and forward are implemented using
Cisco IOS gateways
Describe how T.38 and pass-throughare supported by H.323, SIP. and MGCP
Explain DTMFrelay, its options, and how DTMF relay is supported in MGCP. H.323. and
SIP env ironments
Audio Clarity
This topic describes audio clarity requirements related to VoIP transmission.
Audio Clarity Factors
* Delay: Time it takes for the signal to propagate from one end
to the other end ofthe conversation
* Jitter Variation in the arrival of voice packets
Fidelity: Audio accuracy or quality
* Echo. Usually due to impedance mismatch
* Packet loss: Loss of packets in the network
Sidetone: Allows speakers to hear their own voice
- Background noise: Low-volume noise heard atthefarend of
the conversation
Due to the nature of IP-based forwarding. VoIP packets arc subject to certain transmission
problems. Conditions that are present in the network may introduce problems such as echo,
jitter, or delav. These problems must be addressed with QoS mechanisms.
Ihe clarity (or cleanliness and crispness) ofthe audio signal is of utmost importance. The
listener must be able to clearly understand the speaker. These factors can affect clarity:
Delay: Delav i^the time between the spoken voice and the arrival ofthe electronically
delivered voice at the far end. Delay results from multiple factors, including distance
(propagation delay), coding, compression, serialization, and buffers.
Jitter: Jitter is variation in the arrival of coded speech packets at the far end of a VoIP
network. I he van ing arrival time ofthe packets can cause gaps in the re-creation and
playback ofthe voice signal. These gaps are undesirable and annoy the listener. Delay is
induced in the network by variation in the routes of individual packets, contention, or
congest ion. fhe variable delay can be partly compensated by using dejittcr buffers.
Fidelity: fidelity is the degree to which a system, or a portion of a system, accurately
reproduces, at its output, the essential characteristics ofthe input signal. Human speech
typically requires a bandwidth from 100 lo 10,000 11/.. although 90 percent of speech
intelligence is contained between 100 and .3000 11/..
Echo: Fcho is a result of electrical impedance mismatches in the transmission path. Fcho is
always present, even in traditional telephony networks, but at a level that cannot be
detected by the human ear. The two components that affect echo are amplitude (loudness of
the echo) and delay (the time between the spoken voice and the echoed sound). Fcho is
controlled using suppressors or cancellers.
Packet loss- Voice packets mav be dropped under various conditions such as an unstable
network, network congestion, or too much variable delay in the network. Lost voice
packets are not recoverable, which results in gaps in the conversation that are percept.blc to
the user.
Sidetone: Sidetone is the purposeful design ofthe telephone that allows the speaker to hear
the spoken audio in the earpiece. Without sidetone, the speaker is left with the impress.on
that the telephone instrument isnot working.
Background noise: Background noise is the low-volume audio that is heard from the far-
end connection. Certain bandwidth-saving technologies, such as voice activity detection
(VAD) can eliminate background noise altogether. When VAD is implemented, the
speaker audio path is open to the listener, while the listener audio path is closed to the
speaker. The effect is often that speakers think that the connection is broken because they
hearnothing from theotherend.
)2010ClScoSystemS.lnc. VoIP Call Legs 2-147
Delay
This subtopic evplains the sources ofdelay in an IP network.
Delay Sources
* All network components contribute todelay
Variable delay creates jitter
Picket Flow
Se'ia izator
-'eia, beri3lizatlon
a, II ''r*- SenaIiz atiDm Delay
Pooler
Serialization Delay
Router Router
, j-
**
1 n-i-
I PjtlH
Buffer
Ov erall orabsolute delay can affect VoIP. The delays can cause entire words in the
conversation tobe cut offand therefore can bevery frustrating.
It is important to understand and account for the predictable delay components in the network
when designing anetwork that transports VoIP packets. All potential delays must be considered
to ensure that overall network performance is acceptable. Overall voice quality is afunction of
many (actors, including the compression algorithm, transmission errors, echo cancellation, and
delay,
Thetwodistinct types of delay are as follows:
Kixed delay: Fixed-delay components are predictable and add directly to overall delav on
the connection.
\ ariablc delays: Variable delays arise from queuing delays in the egress trunk buffers that
are located onthe serial port that is connected lothe WAN. These butlers create variable
delays, called jitter, across the network.
The figure illustrates the different types of delav.
Implemerlmg CiscoVoice Communications and QoS(CVOICE) v80
Delay Types
This subtopic describes the characteristics of various delay types.
Delay Types
DSP delay
Processing
delay
Queuing
delay
Time required [orsampling, encoding.
packeSialion, andtne reverse process
Tnetime thatiltakeslora router to takethe
packet from aninput Interfece, examine il,
and put it Intothe output queue
TheSme thata packet mstdes inthe output
queue of a router
NO
(Toted delay)
Yes
(variable delay)
Yes
(variabledelay)
Senal.zat.on Thetime that itlakes totransmit the"bits on No
(fixed delay)
delay thewire
propagation The fame thattakes to propagate apacKet No
delav fromoneenddthelinktoanother (fixed delay)
Buffer to compensate fordelayvariation
delay
Oejtter
buffer
No
(fixed delay)
Orlginatinaand
iefminating gateways
Intermediate routers
Originating gateway
and irrtermerjgts
routers
Originating gateway
and Intermediate
routers
Originating, gateway
andmiarmediate
routers
Terminating gateway
The table describes the delay types, specifies whether the specific delay type adds to the jitter,
and lists the devices that contribute to the defined delay type.
Acceptable Delay (G.114)
This subtopic defines ihe delav -related requirement for transmission of VoIP packets.
Acceptable Delay (G.11
Delay in ms
(onewavi
0-150
150-400
Above 400
Accepiable farmostuserapplications
Acceptable, provided thai administrators areaware of
Ihetransmission timeand itsimpact onIhe
transmission quality ofuser applications
Unacceptable for general network planning purposes
(However, itis recognized (hat in some exceptional
cases, this limit wil be exceeded.)
Values describe acceptable one-way delay
The ITU-T Recommendation G.I 14 specifies acceptable nclvvork delav for voice applications
Ihis recommendation delines three bands of one-way delay, as shown in the table in the figure.
Note
This recommendation is for connections with echo that are adequately controlled implying
that echo cancellers are used Echo cancellers are required when one-way delay exceeds
25 ms fITU-T Recommendation G 1311.
his recommendation is designed for national telecommunications administrations and
therefore, is more stringent than recommendations (hat would normallv be applied in private
voice networks. When the location and business needs ofend users are well known to a
network designer, more delay mav prove acceptable. For private networks, a200-ms one-wav
delay is areasonable goal and a250-ms one-way delay is the limit. The 0.1 14 recommendation
is tor one-way delav onlv and does not account for round-trip delay. Defining the delay as one-
wav is significant, because in asv mmetrical connectivity scenarios (for example with
asymmetric DSL [ADSLj uplinks), avery fast one-way transmission cannot compensate for a
long delav ,n the other direction. The perceived audio quality would deteriorate due to ihe poor
one-wav transmission speed.
Calculating Delay Budget
Network design engineers must consider both variable and fixed delays Variable delays
include oueuing and network delays, while fixed delays include coding, packetization
ser.alizat.on. and dejitter buffer delays. The Delay Budget Calculations table is an example ol
calculating delav budget.
2010 Cisco Systems Inc
Delay Budget Calculations
Delay Type
Fixed (ms) Variable (ms)
Coder delay (digital
signal processor [DSP])
18
Packetization delay
(DSP)
30
Queuing and buffering
8
Serialization (320 kb/s) 1
Network delay -
65
Dejitter buffer 45 -
Totals
98 73
Jitter
Example
This subtopic explains jitter andprovides information on iheacceptable jitter threshold.
One-way jitter should be less than 30 milliseconds.
Exact value depends on codec type
Codecs use different mechanisms to compensate jitter
Steady Stream of Packets
Jitter ts ll-ie
variance in
Time
Same Packet Stream wilh Jitter
.litter is delinedas a variation inthe arrival (delay)of received packets. On the sending side,
packets arc sent in a continuous stream, with the packets spaced evenly. Networkcongestion,
improper queuing, or configurationerrors can cause this steady streamto become uneven,
because the delay between each packet varies instead of remainingconstant, as displayed in the
figure.
When a router receives an audio stream for VoIP, it must compensate for the jitter that is
encountered. The mechanism that manages this function is the playout delay buffer, ordejitter
buffer. The play out delay buffer musl buffer these packets and then play themout in a steady
stream to the DSPs lo be converted back to an analog audio stream. The playout delay buffer,
however, affects overall absolute delay.
TTie limit of acceptable jitter is 30 milliseconds measured in one-way direction. This figure is
an approximate value and depends on the actual codec used for media transmission. Some
codecs perform better than others when compensating jitter.
When a conversation is subjected tojitter, the results can be clearly heard. If the talker says.
"Watson, come here. 1want you." the listener might hear. "Wat....s...on come here.
I wa nt y on." The variable arrival of the packets at the receiving end causes
the speech to be delayed and garbled.
2-152 Implemenling Cisco Voice Communications and QoS (CVOICE) v8 I 2010 Cisco Systems. Inc.
packet loss limit.
Packet Loss
. Most codecs survive the loss of individual packets well.
. The loss of multiple packets in arow deteriorates audio quality.
. in gewial, acceptable packet loss Is 1%.
. The exact value depends on codec type and number of los.packets
in a row
. Codecs implement vanous methods to survive packet loss.
Losl Audio
Packel 1
Lost Packet 2
Packel 3
listener experiences gaps.
Tte accede level for U* c-*-. ^"Î^dlIS
packet loss.
If. conversation experiences packet loss, the effeel is immeâtdy heard If the taiker say,
-Watson, come here. 1want you" the listener might hear, Wat--, come here, >o
Example
VoIPCall Legs 2-153
2-154
Bandwidth Requirements
Thissubtopic defines the bandw idth requirements f
(andwidth Requirements
Values do not include Layer 2overhead
Differs on a hop-by-hop basis
Type of traffic
Bandwidth
Requirement
Voice media 17.106 kb/s Per^ ^^.^^ ^
overhead. Depends onthecodec
type. Thebandwidth should be
guaranteed end-to-end
Scaling (SCCP) 150 b/s Approximate value per ^.irregular
bandwidth usage
or VoIP transmission.
upper lcr l,eadSs b,docs u In ^de , T^T'"" "K '^^^d b>' L^' 3^
subject c,pc,i,in ilh oîf ShM"d be s"^ -"-'-end and no,
The Skinny Client Control Protocol (SCCP) siemlin* mil;
of banduidth per call. The banduidt lagc vcr[ T" W""*** 15" ^
Implementing Cisco Vo.ce Communications and QoS (CVOICE) v8 0
'2010 Cisco Systems. Inc
QoS Requirements
This topic describes the requirements for QoS to ensure proper VoIP transmission.
QoS Mechanisms for VofP
Header compression
Lowers the packet overhead
FRF. 12
Splits large data packets into smaller fragments to allow timely voice
transmission
IP RTP Priority and Frame Relay IP RTP Priority
Prioritization of voice media traffic
Low Latency Queuing (LLQ)
Voice prioritization overdala packets
Multilink PPP (MLP)
Unkag gregation
Resource Reservation Protocol (RSVP)
Method for Cal Admission Control
PSTN fal back
Back up over PSTN if network service below required level
Following are a fewofthe Cisco IOS features that address the requirements of end-to-end QoS
and service differentiation for voice packet delivery;
Header compression: Used in conjunction with Real-Time Transport Protocol (RTP) and
TCP. compresses the extensive RTP or TCP header, resulting in decreased consumption of
available bandwidth for voice traffic. A corresponding reduction in delay is realized.
Frame Relay Fragmentation Implementation Agreement (FRF.12): Ensures
predictability for voice traffic, aiming to provide better throughput on low-speed Krame
Relay links by interleaving delay-sensitive voice traffic on one virtual circuit (VC) with
fragments of a long frame on another VC utilizing the same interface.
IP RTP Priority and Frame Relay IP RTP Priority: Provides a strict priority queuing
scheme that allows delay-sensitive data such as voice to be dequeued and sent before
packets in other queues are dequeued. These features are especially useful on slow-speed
WAN links, including Frame Relay, Multilink PPP (MLP), and Tl ATM links. It works
with weighted fair queuing (WFQ) and class-based WFQ (CBWFQ).
Low Latency Queuing (LLQ): Provides strict priority queuing on ATM VCs and serial
interfaces. This feature allows you to configure the priority status for a class within
CBWFQ and is not limited to User Datagram Protocol (UDP) port numbers, as is IP RTP
Priority.
Multilink PPP(MLP): Allows large packets to be niultilink-encapsulatedand fragmented
so that they are small enough to satisfy the delay requirements of real-time traffic, MLP
also provides a special transmit queue for the smaller, delay-sensitive packets, enabling
them to be sent earlier than other flows.
Resource Reservation Protocol (RSVP): Supports the reservation of resources across an
IP network, allowing end svstems to request QoS guarantees from the network, for
networks supporting VoIP. RSVPin conjunction with features that provide queuing,
traffic shaping, and voice call signalingcan provide Call Admission Control (CAC) for
voice traffic. Cisco also prov ides RSVP support for LLQ and Frame Relay.
Puhlic switched telephone network (PS'FN) fallback: Provides a mechanism to monitor
congestion in the IP network and either redirect calls to the PSTN or reject calls based on
the network congestion.
2-156 Implemenling Cisco VoiceCommunications and QoS (CVOICE] u8 0 2010 Cisco Systems. Inc
QoS Objectives
This subtopic explains the objectives of QoS.
QoS Objectives
Support dedicated bandwidth
Improve loss characteristics
Avoid and manage network congestion
Shape network traffic
Set traffic priorities across the network
To ensure that VoIP is an acceptable replacement for standard PS'I'N telephony services,
customers must receive the same consistently high quality of voice transmission that they
receive with basic telephone services.
In particular. QoS features provide improved and more predictable network operations by
implementing the following objectives:
Support guaranteed bandwidth: Designing the network such that the necessary
bandwidth is always available to support voice and data traffic
Improve loss characteristics: Designing the Frame Relay network such that discard
eligibility is not a factor for frames containing voice, keeping voice below the committed
infonnation rate (CIR)
Avoid and manage network congestion: Ensuring that the LAN and WAN infrastructure
can support the volume of data traffic and voice calls
Shape network traffic: Using traffic-shaping tools to ensure smooth and consistent
deliveiy of frames to the WAN
Set traffic priorities across the network: Marking the voice traffic as priority and
queuing il first
2010 Cisco Systems, Inc. VoIP Call Legs 2-157
Transporting Modulated Data over IP Networks
This topic describes the challenge of transporting fax and modem data over the IP network.
Transporting iVioduiated
Fax and modem traffic consists of digital data modulated into
high-frequency tones.
In contrast to voice, packet loss is much more critical for fax
and modem communications
VoIP compression algorithms are designed for voice, not for
fax or modem data frequencies.
Methods to transmit fax and modem over IP networks:
Terminating modulated signal and transmitting it as data
(fax relay)
Sending the data in-band into the RTP stream
(fax pass-through)
Receiving and converting faxes to files using T 37
(store-and-forward)
An IP. or packel-sw itched, network enables data lo be sent in packets to remote locations. The
data is assembled bv a packet assembler/disassembler (PAD) into individual packets of data,
involving a process of segmentation or subdivision of larger sets of dala as specified by the
native protocol ofthe sending device. Lach packet has a unique identifier that makes it
independent and has its own destination address, liecause the packet is unique and independent,
it can tra\ erse the network in a stream of packets and use different routes. This fact has some
implications for fax transmissions that use data packets instead of an analog signal over a
circuit-switched network.
Differences from Fax Transmission in the PSTN
In IP networks, individual packets that are pari ofthe same data transmission may follow
different physical paths of van ing lengths. They can also experience varying levels of
propagation delav (latency) and delay that is caused by packets being held in packet buffers
awaiting the availability of a subsequent circuit. The packets can also arrive in an order
different from the order in which they entered the network. Ihe destination nod; ofthe network
uses the identifiers and addresses in the packel sequencing information to reassemble the
packets into the correct sequence.
Fax transmissions are designed to operate across a 64-kb/s pulse code modulation (PCM)-
encoded voicc circuit. Ilowever. in packet nctworks. the 64-kb/s stream is often compressed
into a much smaller data rate bv passing it through a DSP. The codecs normally used lo
compress a voice stream in a DSP are designed to compress and decompress human speech, not
fax or modem tones, for this reason, faxes and modems are rarely used in a Vol P network
without some kind of relay or pass-through mechanism in place.
2-158 Implementing Cisco Voice Communicalions and QoS (CVOICE) v8 0
Fax Services over IP Networks
There are three conceptual methods ofcarrying virtual real-time fax machine-to-fax machine
communications across packet networks:
Fax relay: The T,30 fax from the PSTN isdemodulated at the originating gateway. The
demodulated fax content is enveloped intopackets, sent overthenetwork, andremodulated
into the T.30 fax at the terminating gateway.
Note Cisco IOS Software supports T 38 and Cisco fax relay (proprietary).
Fax pass-through: Modulated fax information from the PSTN ispassed in-band end-to-
endover a voice speech pathinan IPnetwork. Thetwopass-through techniques areas
follows:
The configured voice codec is used for thefax transmission. Thistechnique works
only when theconfigured codec is G.7II withnoVAD andnoechocancellation, or
when theconfigured codec is a clear-channel (G.Clear) codec or G.726 or G.732.
I .ow bit-rate codecs cannot be used for fax transmissions.
The gateway dynamically changes the codec from thecodec configured for voice to
G.711 with no VAD and no echo cancellation for the duration ofthe fax session.
Thismethod is specifically referred toas codec "upspeed" or fax pass-through with
upspeed.
Store-and-forward fax: Breaks the fax process into distinct sending and receiving
processes andallows fax messages to bestored between those processes. Store-and-forward
fax is based on the ITU-T T.37 standard, and it also enables fax transmissions to be
received from or delivered to computers rather than fax machines.
)2010 Cisco Systems. Inc. VoIPCall Legs 2-159
Understanding FAX/Modem Pass-Through, Relay,
and Store and Forward
fhis subtopic compares the kev features of pass-through andrelay techniques lor fax and
modem transport.
Pass-Through and Refay
Overview
Availableforfax or modem Available for fax or modem
Carries signal in-band. as RTP packets Carries signal out-of-band as Simple
Packet Relay Transport (SPRT)
packets, over UDP
If supported by both ends, switches lo Ifsupported by both ends, may apply
G.711 or clear channel, no VAD, no echo compression
canceller
Capability of the other end learned via Capability of the other end learned via
Network Signaing Events (NSE), upon Network Signaling Events (NSE), upon
fax or modem signal detection fax or modem signal defection
If not supported by tne other end, falls If not supported by the other end, can
back to regular RTP with configured fall back to pass-through, or from T 38
codec fax relay to Cisco fax relay
This table provides an overview and comparison ofthe key features employed by the pass-
through and relav methods.
fhis topic focuses on the features available to overcome the issues that are involved with
carrying fax and modem signals across an IP network as follows:
Fax and modem pass-through
Fax and modem relav
Fax store and forward
2-160 Implemenling Cisco Voice Communications and QoS (CVOICE) v8 0 2010 Cisco Systems, Inc.
Pass-Through Topology
This subtopic provides an overview ofthe pass-through topology.
Pass-Through Topology
G 71164-Ktts
Encodng
End-to-End CowIbcBofi
With fax pass-through, modulated fax information from the PSTN is passed in-band over a
voice speech path in an IP network.
Fax pass-through is the simplest technique for sending fax over IP networks, but it is not the
default, nor is it the most desirable method of supporting fax over IP. T.38 fax relay provides a
more reliable and error-free method of sending faxes over an IP network. However, some third-
party H.323 and Session Initiation Protocol (SIP) implementations do not support T.38 fax
relay. These same implementations often support fax pass-through.
Fax pass-through is also known as voiceband data by the ITU. Voiceband data refers to the
transport of fax or modemsignals over a voice channel through a packet network with an
encoding appropriate for fax or modem signals. The minimum set of coders for voiceband data
mode is G.711 mu-law and a-law with VAD disabled.
Fax pass-through takes place when incoming T.30 fax data is not demodulated or compressed
for its transit through the packet network. The two endpoints (fax machines or modems)
communicate directly to each other over a transparent IP connection. The gateway does not
distinguish fax calls from voice calls.
On detection of a fax tone on an established VoIP call, the gateways switch into fax pass-
through mode by suspending the voice codec and loading the pass-through parameters for the
duration ofthe fax session. This process, called upspecding, changes the bandwidth that is
needed for the call to the equivalent of G.711.
12010 Cisco Systems. Inc. VoIP Call Legs 2-161
With pass-through, the faxtraffic is carried between the twogateways in RTPpackets usingan
uncompressed fonnat resembling the G.711 codec. This method of transporting fax traffic takes
a constant 64-kb/s (pavload) stream plus its IP overhead end-to-end for the duration ofthe call.
IP overhead is 16kb/s for normal voice traflic. but when using fax pass-through, the
packetization period is reduced from 20 ms to 10 ms. which means that hall'as much data can
be put into each frame. The result is that vou need twice us many frames and twice as much IP
overhead. For pass-through, the total bandwidth is 64 plus 32 kb/s, for a total of 96 kb/s. For
normal voice traffic, total bandwidth is 64 plus 16 kb/s. for a total of 80 kb/s. The table
compares a G.711 VoIP call that uses 20-ms packetization with a G.711 fax pass-through call
that uses 10-ms packetization.
Packetization G.711 Payload Overhead for
Layers 3 and 4
Packet Size Bit Rate
10 ms 80 byte 40 byte 120 byte 96 kb/s
20 ms 160 byte 40 byte 200 byte 80 kb/s
Packet redundancv mav be used to mitigate the effects of packet loss in the IP network. Fven
so. fax pass-through remains susceptible to packet loss, jitter, and latency in the IP network,
fhe two endpoints must be clocked synchronously fortius type oflransporl to work
predictably.
Performance mav become an issue. To attempt lo mitigate packet loss in the network,
redundant encoding (l.v or one repeat ofthe original packet) is used, which doubles the amount
of data transferred in each packet. The doubling of packets imposes a limitation on the total
number of ports that can run fax pass-through at one time. One fax pass-through session with
redundancv needs as much bandwidth as two G.7I I calls without VAD.
2-162 Implementing Cisco Voice Communications and QoS (CVOICE) v8 0 ) 2OI0 Cisco Systems. Inc
Pass-Through
This subtopic explains the pass-lhrough transportof fax and modem data.
Fax
Pass-Through
Fax
Works only when the configured codec is G.711 or clear channel.
Some gateways have limited port numbers for simultaneous use.
- VAD and echo cancellation are disabled.
May issue redundant packets.
Supported under the following call control protocols:
- H.323
- SIP
- MGCP
The following factors determine the deployment of fax pass-through:
Fax pass-through does not support the switch from G.Clear to G.711. If fax pass-through
and the G.Clear codec are both configured, the gateway cannot detect the fax tone.
Fax pass-through is the state ofthe channel after the fax upspeed process has occurred. In
fax pass-through mode, gateways do not distinguish a fax call from a voice call, fax
communication between the two fax machines is carried in its entirety in-band over a voice
call. When using fax pass-through with upspeed, the gateways are to some extent aware of
the fax call. Although relay mechanisms are not employed, with upspeed, the gateways
recognize a called terminal identification fax tone, automatically change the voice codec to
G.711 if necessary' (thus the designation upspeed), and turn off echo cancellation and VAD
for the duration ofthe call, fax pass-through disables compression, echo cancellation, and
issues redundant packets to ensure complete transmission.
Fax pass-through is supported under the following call control protocols:
H.323
SIP
Media Gateway Control Protocol (MGCP)
Modem
Pass-Through
Modem
* Works only when the configured codec is G.711 or clear-channel.
VAD and echo cancellation need to be disabled.
* Modem pass-through over VoIP performs these functions:
- Represses processing functions
May issues redundant packets
Provides static jitter buffers
Differentiates modem signals from voice and fax signals
Reliably maintains a modem connection across the packet
network
Modem pass-through provides the transport of modem signals through a packet network by
using an uncompressed voice codec: G.711 mu-law or a-lavv. It is based on the same logic as
fax pass-through: An analog voice stream is encoded into G.711, passed through the network,
and decoded back to analog signals at the far end. Although modem pass-through remains
susceptible to packet loss, jitter, and latency in the 1Pnetwork, packet redundancy may be used
to mitigate the efleets of packet loss in the IP network.
fhe following factors determine the deployment of modem pass-through:
Modem pass-lhrough does not support the switch from G.Clear lo G.711.
VAD and echo cancellation are disabled when modem pass-through is used.
Modem pass-through over VoIP performs the following functions:
Represses processing functions such as compression, echo cancellation, high-pass
filter, and VAD
Issues redundant packets to protect against random packet drops
I'rov ides sialic jitter buffers of 200 ms to protect against clock skew
Discriminates modem signals from voice and fax signals, indicating ihe detection of
the modem signal across the connection and placing the connection in a slate thai
transports the signal across the network with (he least amount of distortion
Reliably maintains a modem connection across the packet network for a long
duration under normal network conditions.
2-164 Implementing Cisco Voice Communicalions and QoS (CVOICE) v8.0 )20I0 Cisco Systems, Inc.
Relay Topology
This subtopic presents thetopology of fax and modem relay.
Relay Topology
Unlike pass-through, relay demodulates the fax bitsat the local gateway, sendsthe information
acrossthe voice network usingthe Simple Packet RelayTransport (SPRT)protocol, and then
remodulates the bits back into tones at the far gateway. SPRT is a protocol running over UDP
packets to theothergateway, where the signal is re-created, remodulated, andpassed tothe
receiv ing device. The faxmachines andmodemson either end are sendingand receiving tones
and are not aware that a demodulation/modulationprocess is occurring on the originating and
tenninating gateways.
Relay
Fax
This subtopic explains the relav transport of fa\ and modem data.
Fax
* Two fax relay modes'
Cisco proprietary
* Falls back to regular RTP
T.38
' Standards-based
* Can fall back to Cisco fax relay, fax pass-through,
orregular RTP
* Both modes supported by H.323/S1P/MGCP
* Fax relay Packet Loss Concealment
Cisco provides the following two methods for fax relay:
Cisco fax relay: A propnetan. Cisco method and the default on most platforms if a fax
method is not explicitly configured. It is an RTP-based transmission method that uses
proprietary signaling and encoding mechanisms. In Ciscofax relay mode, gateways
terminate 1.30faxsignaling by spoofing a virtual fax machine to the locally attached fax
machine, fhe mechanism for Ciscofaxrelav is thesame for calls that are controlled by
SIP. MGCP, and H.323 call control protocols,
T.38 fax relay: Based on the 1IIJ-T standard. T.38 is a real-time fax transmission method
that allows two fax machines to communicate as if therewere a direel phon^linebetween
them, "fheT.38 fax relay feature can be conllgurcd for 11,323. SIP, and MGCPcall-control
protocols.
2-166 Implementing Cisco Voice Communicalions and QoS (CVOICE) v8.0 2010 Cisco Systems. Inc
Modem
Relay
Modem
Cisco proprietary method
Can fall back to modem pass-through
Provides these features:
- Modem tone detection and signaling
Optional payload redundancy
Dynamic and static jitter buffers
Modem relay is a Ciscoproprietary method for demodulating, transporting, and remodulating
modem data for transport over an IP network. On detection ofthe modem answertone, the
gateways switchintomodem pass-through modeand then, if the call menusignal is detected,
the two gateways switch into modemrelay mode. In this implementation, the call starts out as a
voice call, and then switches into modempass-through mode, and then into modem relay mode.
Modem relaysignificantly reducesthe effects that droppedpackets, latency,andjitter have on
the modemsession. Compared to modem pass-through, it consumes less bandwidth. If the
gateways fail lo agreeon the modem relay, the transmission falls backto modem pass-through.
Modem relay includes the following features:
Modem tone detection and signaling: Modem relay supports V.34 modulation and the
V.42 error correction and link-layer protocol with maximum transfer rates of up to 33.6
kb/s. It forces higher-rate modems to train down lo the supported rates. Signaling support
includes SIP, MGCP or SGCP, and H.323. The gateways negotiate whether to use the
modem relay mode, the gateway exchange identification (XID), and payload type for
Named Signaling Event (NSE) packets.
Optional payload redundancy: Payload redundancy causes the gateway to send redundant
packets. Redundancycan be enabled in one or both ofthe gateways. When only a single
gateway is configured for redundancy, the other gateway receives the packets correctly but
does not produce redundant packets. When redundancy is enabled, 10-mssample-sized
packets are sent. When redundancy is disabled, 20-ms sample-sized packets are sent.
Note By default, the modem relay and redundancy are disabled.
) 2010 Cisco Systems, Inc VoIP Call Legs 2-167
Store-and-Forward Fax
This subtopic explains the concept of fax store-and-forward.
2-i(
>re-and~Forwarc
On-ramp receives faxes that are delivered as email attachments.
PSTN
Off-ramp sends standard email messages that are delivered
as faxes.
PSTN
In store-and-forward fax. the transmitting gateway is relerred to as an "on-ramp gateway," and
the tenninating gatewav is referred to as an "off-ramp gateway." Here is how they work:
On-ramp faxing: A voice gateway that manages incoming calls Irom a standard fax
machine or the PSTN converts a traditional Group 3 (G3) fax to an email message with a
lagged Image File Fonnat (I IFF) attachment. 'Flic fax email message and attachment are
managed by an email server while traversing the packet network and can he stored for later
deliverv or delivered immediately to a PC or to an off-ramp gateway.
Off-ramp faxing: A voice gateway that manages calls going out from the network to a fax
machine or the PSTN converts a fax email with a TIKI" attachment into a traditional fax
fonnal thai can be delivered to a standard fax machine or the PSTN.
On-ramp and off-ramp faxing processes can be combined on a single gateway, or they can
occur on separate gatewavs. Store-and-forward fax uses two different interactive voice
response (IVR) applications for on-ramp and off-ramp functionality. The applications are
implemented in two Tool Command Language (Tel) scripts that you can download from
http://www.cisco.com.
The basic functionalilv of store-and-forward fax is facilitated through Simple Mail Transfer
Protocol (SM I P). vsith additional functionalilv that provides confirmation of delivery using
existing SMTP mechanisms, such as Extended SMTP (FSMTP).
Implementing Cisco Voice Communicalions and QoS (CVOICE| v8 0 ) 2010 Cisco Systems. Inc
hv
Gateway Signaling Protocols, and Fax and
Modem Pass-Through and Relay
This topic describes thesignaling procedures for fax and modem pass-through and relay.
Fax and Modem Pass-Through
G3Fax Original
Initiates the Coll Gateway
Change Codec
Vlip r-a"
Call Control Issues NSE
NSE Accept
Change Codec
i
The ligure illustrates a fax and modempass-through operation. When a terminating gateway-
detects a called terminal identification (CED) lone from a called fax machine, the tenninating
gateway exchanges the voice codec that was negotiated during the voice call setup for a G.711
codec and turns off echo cancellation and VAD. This switchover is communicated to the
originating gateway, which allows the fax machines to transfer modemsignals as though they
were traversing the PSTN, If the voice codec that was configured and negotiated for the VoIP
call is G.711 when the CED tone is detected, there is no need to make any changes to the
session other than turning off echo cancellation and VAD.
If pass-through is supported, the following events occur:
1. For the duration ofthe call, the DSP listens for the 2100-H/ CED tone to detect a fax or
modem on ihe line.
2. If the CED tone is heard, an internal event is generated to alert the call control stack that a
fax or modem changeover is required.
3. The call control stack on the originating gateway instructs the DSP to send an NSE to the
terminating gateway, informing the terminating gateway ofthe request to carry out a codec
change.
4. If the terminating gateway supports NSEs, it responds to the originating gateway-
instruction and loads the new codec. The fax machines are able to communicate on an end-
to-end basis with no further intervention by the voice gateways.
i 2010 Cisco Systems. Inc. VoIP Call Legs 2-169
Control of fax pass-through is achieved through NSEs that are sent in the RTPstream. NSFs
are a proprietary Cisco version of lETF-standard named telephony events (NTFs), which are
specially marked data packets used to digitally convey telephony signaling tones and events.
NSFs use different event values than NTFs and are generally sent with RTP payload type 100.
whereas NTFs use pavload type 101. NSEs and NTFs provide a more reliable way to
communicate tones and events bv using a single packet rather than a series of in-band packets
that can be corrupted or partially lost.
Fax pass-through and fax pass-through with upspeed use peer-to-peer NSEs within the RTP
stream or bearer stream to coordinate codec sw itchover and to disable echo cancellation and
VAD. Redundant packets can be senl lo improve reliability when the probability of packet loss
is high.
When a DSP is put into voice mode at the beginning of a VoIP call, the DSP is informed by the
call control stack whether the control protocol can support pass-through or not.
2-170 Implementing Cisco VoiceCommunications and QoS (CVOICE) v8.0 2010 Cisco Systems. Inc
*->
Cisco Fax Relay
This subtopic explains the Cisco proprietary fax relay functionality.
Cisco Fax Relay
G3Fax Gateway
Initiates the Call
CED Tone
DIS Message
JP Network
Fax Relay Switchover (PT96)
Send Codec ACK (PT97)
Gateway
Download Codec Download Codec
Codec Download Done (PT96)
Codec Downioaa ACK (PT97)
Fax Relay Established
The figure illustrates aCisco fax relay operation. When a DSP isput into voice mode atthe
beginning of a VoIP call, the DSP isinformed bythecall control stack whether fax relay is
supported and. if it issupported, whether it isCisco fax relay orT.38 fax relay. IfCisco fax
relay is supported, the followingevents occur:
Step1 Initially, a VoIP call is established as if it werea normal speech call. Call control
procedures arefollowed and theDSP is put into voice mode, affer which human
speechis expectedto be receivedand processed.
Step 2 Atany time during the lifeofthe call, if a fax answer orcalling tone (Answer back
[ANSam] or CED) isheard, the DSP doesnot interfere with the speech processing.
The ANSamor CED tone causes a switch to modem pass-through, if enabled, to
allowthe tone to pass cleanly to the remote fax.
Step3 Anormal fax machine, aftergenerating a CEDor hearing a calling (CNG) tone,
sendsa digital information signal (DIS)messagewiththe capabilities ofthe fax
machine. The DSP in the Cisco IOS gateway attached to the fax machine that
generated the DIS message (nonnally thetenninating gateway) detects the High-
Level Data Link Control (HDLC) flag sequence at the start ofthe DIS message and
initiates faxrelayswitchover. The DSPalso triggersan internal event to notify the
call control stack that fax switchover is required. The call control stack then instructs
the DSPto changethe RTP payloadtype to 96 and to sendthis payloadtype to the
originating gateway.
Step 4 When the DSP on the originating gateway receives an RTP packet with the payload
tvpe set to 96. it triggers an event to inform its own call control stack that a fax
changeover has been requested by the remote gateway. The originating gateway then
sends an RTP packet totheterminating gateway with payload type 97to indicate
that the originating gateway has started the fax changeover. When the tenninating
gatewav receives thepayload type 97 packet, thepacket serves as an
acknowledgment, fhe terminating gateway starts the fax codecdownload and is
rcadv for fax relay.
Step 5 Once the originating gatewav has completed the codec download, il sends R'l'P
packets with pavload tvpe 96totheterminating gateway. Fhe tenninating gateway
responds w1th anRTP packet with pav load type 97. and fax relay can begin between
the two gatewavs. As part ofthe fax codecdownload, other parameters suchas
VAD. jitter buffers, and echo cancellation are changed to suit the different
characteristics of a fax call.
During faxrelav operation, the 1.30 analog fax signalsarc received from the PSTN or from a
directlv attached fax machine. The 1.30 fax signals aredemodulated by a DSP onthe gateway
and then packetized and sent across the VoIP network asdata. The terminating gateway
decodes the data stream and remodulates the T..10 analogfax signalsto be sent to the PSTN or
to a destination fax machine.
The messages that aredemodulated and rcmodulated arepredominantly the phase B, phase D,
and phase Emessages of a T.30 transaction. Most ofthe messages are passed across without
am interference, but certain messages are modified according lo the constraints ofthe VoIP
network.
During phase B. each fax machine interrogates thecapabilities ofthe other. They expect to
communicate with eachother across a 64-kb/s PSTN circuit, and they attempt lo make best use
ofthe available bandwidth and circuit qualitv of a 64-kb/svoicepath. However, in a VoIP
network, the fax machines do not have a 64-kb/s PSTN circuit available. Thebandwidth per
call is probably less than 64 kb/s. and the circuit is not considered a clear circuit.
Because transmission paths in Vol!1 networks arc more limited than in the PSTN, the
Cisco IOS command-line interface (CL1) is used to adjust fax settings onthe VoIP dial peer.
The adjusted fax settings restrict the facilities that are available lo fax machines across the VoIP
call legand are also usedto modifv values in DIS and NSFmessages that arc received from fax
machines.
2-172 Implementing Cisco Voice Communications and OoS (CVOICF) v8 0 2010 Cisco Systems, Inc
T.38 Fax Relay
This subtopic explains the T.38 fax relay standard.
H.323
T.38 Fax Relay
H.323
G3Fax
initiates thecal
T30 VoIP Call T.30
CED Tone
Mode Requesl
DIS Message
Mode Requesl ACK
Close VoIP and Open T38 Channels
T.38 UDP Packets
1
The figure illustrates an 11.323 T.38 relay operation. The T.38 fax relay feature provides an
ITU-T standards-based method andprotocols for fax relay. Dataispaekctized andencapsulated
according to the T.38 standard. The encoding ofthe packet headers and themechanism to
switch from VoIPmodeto fax relaymodeare clearlydefinedinthe specification. Annexes to
thebasic specification include details foroperation under SIP and H.323 call control protocols.
The H.245 messages are exchanged in the following sequence:
Step 1 Initially, a VoIP call isestablished as if it were a normal speech call. Call control
procedures are followed andthe DSP is put intovoice mode, afterwhich human
speechis expectedto be receivedand processed.
Step 2 At any time duringthe life of the call, if a fax answeror callingtone (ANSam or
CKD) is heard, the DSPdoes not interfere with the speech processing. The ANSam
or CLD tone causes a switch to modem pass-through, if enabled, to allow the tone to
pass cleanly to the remote fax.
Step 3 A normal fax machine, after generating a CED or hearing a CNG, sends a DIS
messagewiththe capabilities ofthe fax machine. The DSPin theCisco IOSgateway
attached to the fax machine that generated the DIS message (normally the
terminating gateway) detectsthe HDLC flagsequenceat the start ofthe DIS
messageand initiates faxrelayswitchover. The DSPalso triggersan internal event
to notify the call control stackthat fax switchover is required. The call controlstack
then instructs the DSPto change the RTP payload type to 96 and to send this
payload type to the originating gateway.
Step 4 The detecting terminating gateway sends a ModeRequest message to the originating
gateway, and the originating gateway responds with a ModeRequestAck.
Step5 'fhe originating gateway sends a closeLogicalChannel message toclose its VoIP
UDP port, andthetenninating gateway responds with a closcLogicalChannelAck
message while it closes the VoIP port.
Step6 fhe originating gatewav sendsan opcnLogicalChannel message that indicates to
which port to send theT.38 UDPinformation on the originating gateway, andthe
tenninatinggatewav responds withan openLogicalChannelAck message.
Step7 Thetenninatinggateway sendsa closeLogicalChannel message to close its VoIP
UDP port, andthe originating gateway responds witha closcLogicalChannelAck
message.
Step8 Ihe terminating gatewav sends an openLogicalChanne! message that indicates to
which port to sendtheT.38 UDP stream, andthe originating gateway responds with
an openl ogicalChannelAck message.
Step 9 1.38-cncoded UDP packets flow back and forth. At the end ofthe fax transmission,
either gatewav can initiate another ModeRequest message to return to VoIP mode.
T.38 fax relay uses data redundancy to accommodate packet loss. During T.38 call
establishment, voice gateways indicate the level of packet redundancy that they incorporate in
their transmissionof fax UDPtransport layer packets. You can configure the level of
redundancv (the numberof timesthat the packet is repeated) on Cisco IOS gateways.
fhe T.38 Annex B standard defines the mechanism that is used to switch over l>om voice mode
to T.38 faxmodeduringa call. The abilityto support T.38must be indicated duringthe initial
VoIP eall setup. If the DSP on the gateway is capable of supporting T.38 mode, this
infonnation is indicated during the 11.245 negotiation procedures as part ofthe regular 11.323
VoIP call setup.
After the VoIP call setup is completed, the DSP continues to listen for a fax tone. When a fax
tone is heard, the DSPsignals the receipt of fax tone to the call control layer, which then
initiates fax changeover as specified in the T.38 Annex B procedures.
Implementing Cisco Voice Communications and QoS (CVOICE) u8 0 2010 Cisco Systems, Inc
SIP
T.38 Fax Relay
SIP
G3Fai
Initiates the Call
DIS Message
VoIP Call T30
INVITE (T3B in SDP)
ZOO OK
ACK
T3BUDP Packets
'
When SIPis the call control protocol, T.38AnnexDprocedures are used for the changeover
from VoIP to fax mode during a call. Initially, a normal VoIP call is established using SIP
INVITE messages. TheDSPneeds to beinformed that it cansupport T.38 mode whileit is put
into voicemode. Then, duringthe call, when the DSPdetectsfax HDLC flags, it signalsthe
detection ofthe flagsto the call control layer, and thecall control layer initiates a SIP INVITF
message, midcall. to signal the desire to changethe mediastream.
The SIP T.38 fax relay call flow is presented in the following list:
Step1 Initially, a VoIP call is established as if it werea normal speech call. Call control
procedures are followed andthe DSP is put intovoice mode, afterwhich human
speech is expected to be received and processed.
Step 2 At any time duringthe life of the call, if a faxansweror callingtone (ANSam or
CLD) is heard, the DSPdoes not interfere with the speech processing, 'fhe ANSam
or CED tone causes a switch to modempass-through, if enabled, to allowthe tone to
pass cleanly to the remote fax.
Step 3 A normal fax machine, after generating a CED or hearing a CNG, sends a DIS
messagewiththe capabilities ofthe fax machine. The DSPin the Cisco IOSgatewav
attached to the fax machine that generated the DIS message (normally the
terminating gateway) detects the HDLCflag sequence at the start ofthe DIS
message and initiates fax relay switchover. The DSP also triggers an internal event
to notify the call controlstackthat fax switchover is required. 'Fhecall control stack
then instructs the DSP to change the RTP payload type lo 96 and to send this
pavload type to the originating gateway.
Step 4 The terminatinggateway delects a fax V.21 flag sequence and sends an INVITE
message with T.38 details in the SDP field to the originating gateway or to the SIP
proxy server, depending on the network topology.
>2010 Cisco Systems. Inc VoIP Call Legs 2-175
Step5 Theoriginating gatewav receives theINVITT: message andsends back a 200OK
message.
Step6 Ihe tenninatinggatewav acknowledges the 200 OKmessage and sends an ACK
message directly to the originating gateway.
Step 7 The originating gatewav starts sending T.38 UDP packets instead of VoIP UDP
packets across the same ports.
Step8 At the end ofthe fax transmission, another INVFfi: message can be sent to return lo
VoIP mode.
MGCP
T.38 Fax Relay
MGCP
MGCP T.38 fax relay provides two modes of
implementation:
Gateway-controlled mode:
- Gateways negotiatefaxrelaytransmission byexchanging
data in SDP messages.
- Allows the use of MGCP-based T.38 fax without the
necessity of upgrading the call agent software.
Call agent-controlled mode:
- Call agents instructgateways to process fax traffic.
- Call agent can instruct gateways to revert to
gateway-controlled mode if itcannot manage fax control.
MGCP T.38 faxrelayprovides two modesof implementation as follows:
Gateway-controlled mode: Gatewaysnegotiate faxrelay transmission by exchanging
capability information inSession Description Protocol (SDP) messages. Transmission of
SDP messages istransparent to thecall agent. Gateway-controlled mode allows the useof
an MGCP-basedT.38 fax without the necessity of upgrading the call agent software to
support the feature.
Call agent-controlled mode: Call agents use MGCP messaging to instructgateways to
process fax traffic. For MGCP T.38 fax relay, call agents canalsoinstruct gateways to
revert togateway-control ledmode if thecall agent is unable tomanage the fax control
messaging traffic, as is the case inoverloaded or congestednetworks.
MGCP-based T.38 faxrelayenables interworking between theT.38 application that exists on
Ciscogateways andthe MGCP applications on call agents.
MGCP-basedT.38 fax relay call flow is presented in the following list:
Step 1 A call is initially established as a voice call.
Step2 The gateways advertise capabilities in an SDPexchange duringconnection
establishment.
Step3 If bothgateways do not supportT.38 fax relay, faxpass-through is used for fax
transmission. If both gateways support T.38, they attempt to switch lo T.38 upon fax
tone detection. The existing audio channel is used for T.38 fax relay, and the
existing connection port is reused lo minimize delay. If failure occurs at some point
during the switch to T.38. the call reverts to the original settings that it had as a
voice eall. If this failure occurs, a fallback to fax pass-through is not supported.
Step 4 Upon completion ofthe fax image transfer, the connection remains established and
reverts to a voice call using the previously designated codec, unless the call agent
instructs the gateways to do otherwise.
)2010 Cisco Systems, Inc. VoIP Call Legs 2-177
Afax relav MGCP event allows the gateway tonotify the call agent of the status (start, stop, or
failure) ofT.38 processing for the connection, fhis event issent in both call agent-controlled
and gatewav-controlled mode.
Gateway-Controlled MGCP T.38 Fax Relay
In gatewav-controlled mode, a call agent uses the fx: extension of Ihe local connection option
(LCO) to instruct a gateway about how lo processa call. Gateways do not need instruction from
the call agent to switch toT.38 mode, fhis mode is used if the call agent has notbeen upgraded
to support T.38and MGCP interworking. or if the call agent does not want lo manage fax calls.
Ciatcw av -controlled modecan also be usedto bv pass the message delayoverheadcausedby-
call agent handling, for example, to meet time requirements for switchover lo T.38 mode. If the
call agentdoes not specif) the modeto thegateway, the gateway defaults lo gateway-controlled
mode.
Ingatewav-controlled mode, the gatewavs exchange NSFs byperforming the following
actions:
Instruct the peer gatewav- to switch to T.38 for a fax transmission.
Fither acknow ledge the switch andthe readiness ofthe gateway to accept T.38packets or
indicatethat the gatewav cannot accept T.38 packets.
Call Agent-Controlled MGCP T.38 Fax Relay
In eall agent-controlled mode, the eall agent can instruct the gateway to switch to T.38 for a
call. Ihe call agent-controlledmode enables 1.38 fax relay interworkingbetween 11.323
gatewav, and MGCP gateways and between two MGCP gateways under the control of a call
agent. Fhis feature supersedes previous methods for call agent-controlled fax relay and
introduces the following gatewa_v capabilities:
Abilitv to accept the MGCP I XRpackage, to receive the fxr prefix in commands fromthe
call agent, and to send the FXRprefix in notifications to the call agent.
Abilitv to accept a new port when the gateway is switching from voice to fax transmission
during a call. 1his new abilitv allows successful T.38 call agent-controlled fax
communications between H.323 and MGCP gateways in those situations in which the
11,323gatewav assigns anew port when it is changing a call from voice to fix.
2-178 Implementing Cisco Voice Communicalions and QoS (CVOICE) vS 0 2010 Cisco Systems. Inc
DTMF Support
This topic describes the DTMF relay technologies and their support by gateway signaling
protocols.
DTMF Support Overview
DTMF tones are distorted with most low-bandwidth codecs,
such as G.729 and G.723.
- Problems when accessing automated DTMF-based systems,
such as voice mail, menu-based automatic call distributor
(ACD), and automated banking
DTMF relay sends DTMF tones withgreater fidelity than is
possible in-band.
DTMF is the tone generated on a touch-lone phonewhenthe keypaddigits are pressed.
Gateways sendthese tones in the RTP stream bydefault. Thisdefault behavior is fine when the
voice streamis sent uncompressed, but problems arise whengateways use compression
algorithms to send voice across slower WANlinks.
During a call. DTMF may be entered to access IVRsystems suchas voice mail andautomated
banking services. Although DTMF is usually transported accurately when high-bit-rate voice
codecssuchas G.711 are beingused, lovv-bit-rate codecssuch as G.729and G.723.1 are highly
optimized for voicepatternsand tendto distort DTMF tones. As a result, IVRsystems may not
correctly recogni/.e the tones.
DTMF relay solves the problemof DTMF distortion by transporting DTMF tones "out of
band." or separate from the RTP voice stream.
>2010Cisco Systems. Inc. VoIP Call Legs 2-179
DTMF Support
H.323
This topic compares the DTMF relay methods supported by each gateway signaling protocol.
H.323
DTMF method
Cisco
proprietary
H 245 signal
H.245
alphanumeric
RTP Named
Telephony
Events (NTEs)
None
Description"
In-band, DTMF tones carried in the same RTP channel as voice.
DTMFtones are encoded oSfferently from voice and are identified
as paybad type 121. Requires Cisco gateways at both ends.
Gut-of-band DTMFsenl through H.245 signaling instead of RTP.
Signals tone length.
Out-of-band. DTMFsent through H.245signalinginstead OfRTP.
The tones are transported h H.245 User Input indication
messages.This method does no! send tone lengBi information.
Support required for H.323v2 compliance.
In-band. RFC-based DTMFtransport in RTP. Special NTE RTP
formats exist for DTMFdigits, hookflash, and other telephony
events. With the NTEmethod, the endpoints perform per-call
negotiation ofthe DTMF relay method.
in-band. DTMF tones are left in the audio stream without any
marking, Defaultseaing.
If multiple methods are supported, selection priority is as shown in the table.
Ciscogatewav s support multiple methods of DTMF relay using11.323. Theyare presented in
the order of their selection priority as follows:
Proprietary Cisco: DIMF tones are sent in the same RIP channel as voice data. However,
the DTMF tones are encoded differently fromthe voice samples and are identified as
pavload tvpe 121, which enables the receiver to identify them as DTMF tones. This method
requires the use of Cisco gateways at both the originating and tenninating endpoints ofthe
H.323 call.
H.245 signal: Separates the DTMFdigits fromthe voice streamand sends themthrough
the H.245 signaling channel instead of through the RIP channel, fhe tones are transported
in H.245 I ser Input Indication messages. The H.245 signaling channel is a reliable
channel, so the packets that transport the DIMF tones are guaranteed lo be delivered. In
contrast to 11.245 alphanumeric. H.245 signal passes along tone length information,
therefore addressing a potential problem with the alphanumeric method, fhis method is
optional on 11,323gatewavs.
11.245alphanumeric: Separates the DTMF digits from the voice stream and sends them
through the H.245 signaling channel instead of through the RIP channel. Fhe tones are
transported in H.245 User Input Indication messages, fhis method docs not send tone
length infonnation.
Note All H 323 version 2-compliant systems are required to support the H.245 alphanumeric
method, while support of the H 245 signal method is optional.
Implementing Cisco Voice Communications and QoS (CVOICE) ve.O )2D10 Cisco Systems, Inc.
NTE: Transports DfMF tones in RTP packets according to Section 3of RFC 2833. RFC
2833 defines formats ofNTE RTP packets that are used to transport DTMF digits,
hookflash and other telephony events between two peer endpoints. With the NTE method,
the endpoints perform per-call negotiation ofthe DTMF relay method. They also negotiate
to determine the payload type value for the NTE RTP packets. User preference tor DTMF
relav tvpes is not supported, and DTMF relay forking is not supported.
. 2010 Cisco Systems, Inc. PCall Legs 2-181
SIP
SIP
DTMF method
SIP Notify
RTP Named
Telephony
Events (NTEs)
None
Out-of-band Forwards DTMF tones using SIP NOTIFY messages,
NOTIFY messagesare exchanged bidireclionally between the
originating andterminating gateways for a DTMF everrt during a
call. SIP NOTIFY messages areadvertised in anINVITE message
tothensmote endonly ifthis typeofDTMF relay is negotiated.
In-oand. RFC-based DTMF transport inRTP. Special NTE RTP
formats exist for DTMF digits, hookflash, andother telephony
events. With the NTE method, the endpoints perform per-call
negotiation of the DTMF relay method.
In-band. DTMF tonesare left in theaudio stream without any
marking. Default setting.
If mult
le methods are supported, selection priorityis as shown in the table.
Cisco gateway* support multiple methods ofDTMF relay using 11.323. They are presented in
theorder of their selection priority as follows:
SIP Notify: SIP gateways can use Notify-based out-of-band DTMF relay. Notify-based
out-of-band DIMF relav sends messages bidirectional!) between the originating and
tenninating gatewavs for a DIMF' event during a call. The originating gateway sends an
INVITE message with a SIP Call-Info header toindicate the use ofthe Notify-based out-of-
band DIME relav. fhe tenninating gateway acknowledges the message with an 18x or 200
Response message, which alsouses the Call-Info header. When a DTMF event occurs, the
gatewav sends a SIP Notify message for that event alter the SIP Invite and 18xor 200
Response messages negotiate the Notify-based out-of-band DTMF relay mechanism. In
response, thegateway expects toreceive a 200OKmessage.
NTEs: Transports DTMF tones inRTP packets formatted totransport DIMFdigits,
hookflash. and other telephony events between two peer endpoints. Ihismethod is based
on the same standard as the NIF-based relay in 11.323.
2-182 Implementing Cisco VoiceCommunicalions and QoS (CVOICE) vB.O
MGCP
DTMF Support
MGCP
\Wmm rVTMC mo.hnrf
Description ^mwrn
Cisco
In-band. DSPson Hie gatewayssendandreceiveDTMF digits in-
proprietary band h the voice RTP stream but codes them differently so that
they can be identifiedby the receiver as DTMF tones.
RTP Named RFC-based in-band method. Transports DTMFtones using NTEs
Signaling Event in RTP packets.
(NSE)
RTP Named
RFC-based in-band method. Twomodes: gateway-controlled and
Telephony cat agent-controled. Ingateway-controlted mode, the gateways
Events (NTE)
negotiateDTMF transmissionby exchangingcapablity infomialion
inSDP messages. That transmissiontotransparent to thecal
agent.Incallagent-controiled mode,callagentsuse MGCP
messaging to instruct gatewaystoprocessDTMF traffic
Out-of-band Sends the tones as signals to Ihe call agent out-of-band over Ihe
control channel.
None
In-band. DTMFtones are left in me audio stream without any
marking. Default setting.
Cisco gateways support multiple methods of DTMF relay using MGCP. They are presented in
the order of their selection priority as follows:
Proprietary Cisco: DSPs onthegateways sendandreceive DTMF digits in-band in the
voice RTP streambut codes themdifferently so they can be identified by the receiver as
DTMF tones.
NSE: Conforms to RFC 2833 toprovide a standardized method of DTMF transport using
NTEs in RTPpackets. RFC2833support is standards-based and allowsgreater
interoperabilitywith other gateways and call agents.
NTE: Provides for two modes of implementation:
Gateway-controlled mode: Ingateway-controlled mode, thegateways negotiate
DTMF transmission by exchanging capability information in SDPmessages. That
transmission is transparent to the call agent. Gateway-controlled modeallowsthe
use ofthe DTMF relay feature withoutupgrading the call agent softwareto support
the feature.
Call agent-controlled mode: In call agent-controlled mode, call agentsuse MGCP
messagingto instruct gateways to process DTMFtraffic.
Out-of-band: Sends the tones as signals to the call agent out-of-band over the control
channel. The call agent interprets the signals and passes themon.
None: No DTMF method is used and DTMF tones are left in the RTP streams without any
marking. This is the default setting.
Summary
Summary
Audioquality is affected by packet delay, jitter, packet loss,
distortion level, echo, sidetone, and background noise.
QoS improves loss characteristics, avoids and manages
network congestion, and prioritizes voice traffic.
Modulated data such as fax or modem is very vulnerable to
codec distortion.
Fax transmission methods include pass-through, relay, and
store-and-forward. Modem transmission uses pass-through,
or relay.
Fax and modem pass-through, as well as Cisco proprietary
and T.38 fax relay are supported by H.323, SIP, and MGCP.
DTMF tones can be carried in-band without any special
handling, in-band with special handling (Cisco proprietary,
RTP-NTE, RTP-NSE), and out-of-band (H.245 signal, H.245
alphanumeric, and SIP-Notify).
2-184 Implementing Cisco Voice Communications and QoS (CVOICE] vB.O 2010 Cisco Systems, Inc.
Lesson 6
Configuring VoIP Call Legs
Overview
Successful implementation of a VoIP network depends upon thecorrect application of dial
peers, thedigits that thedial peers match, and theservices thatthey specify. You must have in-
depth knowledge of dial-peer configuration options andtheiruses. Thislesson discusses the
configuration of VoIP dial peers.
Objectives
Upon completing this lesson, youwill be abletodescribe howto configure VoIP call legsina
gateway. This abilityincludes beingable to meet these objectives:
Explain the keyconfiguration components of a basic VoIPdial peer and describe howto
configure VoIP dial peers
Describe how to configure DTMF relay
Explainhow to configure fax/modempass-through and relay
Describe howto configure a single codec or codec negotiation on an SIP and H.323
gateway
Explain howto limit the numberof concurrent calls on a VoIPdial peer
Configuration Components of VoIP Dial Peer
Ihis topic describes the configuration components ofa VoIP dial peer and provides VoIP dial
peer characteristics.
incoming called-numbe
direct-inward-dial
incoming called-nijikber
port 1/0/O
IP WAN
incoming e ailed-number .
direct-inward-dial
dial-peer vice 2 vo
'P
destinatio n-pattern 2001
port 1/0/0
dial-peer v 3ice 2002 pots
deatinatio n-pattern 2002
port 1/0/1
dial-peer v sice 1000 voip
destinatio i-pattarn 100.
session ta rget ipv4 10.1.1.1
The configuration examplein the figure illuslrates the key components of VoIPdial-peer
configuration, fhe second dial peer on each gateway is used lo malch incoming VoIP calls, fhe
VoIP dial peers 2000 and 1000 are configured to forward calls to the remote location,
respectively . For call forwarding, the VoIP dial peer uses the destination-pattern and the
session target commands. For matching inbound VoIP dial peers, the priority of matching is
defined in this order: incoming called-number. answer-address, and destination-pattern.
Implementing Cisco Voice Communications and QoS (CVOICE) v8 i >20IOCisco Systems, Inc
VoIP Dial Peer Characteristics
This subtopic describes the characteristics ofVoIP dial peers.
VoIP Dial Peer Characteristics
Signalingprotocol
Source IP
address
Digit consumption
Session targel
Inbound dial peer
matching
Outbound dial
peer matching
Direct Inward Dial
Default H.323. Canbe changedwith Session protocol Slpv2
command.
Outgoing interface address. Canbe changed with h323-gateway
voipbind sreaddr (interface mode)or bind(sipmode)
command.
Nodigits areconsumed (equivalent toforward-digits all).
IPaddress, DNSname, gatekeeper (RAS), or SIP server.
Configured with session target command.
incoming caUed-number, answer-address, destination-pattern.
Default dial peer: anycodec, noDTMF relay, IPprecedence 0.
VAD enabled.
Most elicit match on destination-pattern.
Doesnot apply. Related only to inbound POTSdialpeers.
Yonshould consider the following aspects when configuring VoIP dial peers:
Signaling protocol: H.323 isthe default setting. The protocol can be changed to Session
Initiation Protocol (SIP) v2. Media Gateway Control Protocol (MGCP) control can only be
configured for plain old telephone service (POTS) dial peers; it isnot available for VoIP.
Source IP address: By default, thesource IP address isdefined bytheIPlayer. The
routing table defines the outgoing interface to reach adefined session target. The outgoing
interface address is used as thesource address for both signaling andmedia. This behavior
can bemodified by interface binding: the h323-gateway voip bindsreaddr command for
H.323 (interface mode), or the bind command forSIP(SIP mode).
Digit consumption: Unlike POTS dial peers, VoIP dial peers do not consume any digits.
Session target: Thetarget ofthe VoIP session can beset toanIPaddress, DNS name,
gatekeeper (RAS). orSIP server. Itisconfigured with the session target command.
Inbound dial peermatching: Performed with these commands inthis order: incoming
called-number, answer-address, anddestination-pattern. If no inbound dial peeris
matched, the default peer istried. The default peer has these parameters: any codec, no dual
tone multifrequency (DTMF) relay, IPprecedence, voice activity detection (VAD) enabled.
Ifthese parameters cannot benegotiated (for example, ifthe originating gateway has VAD
disabled), the call fails.
Outbound dial peermatching: Themostexplicit match ofthe destination-pattern
command.
Direct inward dialing (DID): Notapplicable to VoIP dial peers; available for inbound
POTS dial peers only.
Configuring DTMF Relay
This topic describes how to configure DTMF relay negotiation in 11.323 and Sll
dtmf-relay {[cisco-rtp] [h245-alphanumeric] [h245-signal]
[rtp-nte [digit-drop]] [sip-notify]}
1H.323 options (in priority order): cisco-rtp, h245-signal,
h245-alphanumeric rtp-nte, none
SIPoptions (in priority order): sip-notify, rtp-nte, none
If no common dtmf-relay method is negotiated, the call is set
up with dtmf-relay none (DTMF tones left in the RTP
channel)
Digit-drop option(available for RTP-NTE) is requiredfor
Cisco Unified Border Element interworking with out-of-band
methods (H.245 and SIP-Notify)
DTMF relay methods for SIP and H.323 areconfigured inIhc dial-peer configuration mode,
using the dtmf-relay command. If this command is not configured, the DTMF tones are
disabled and sent in-band. fhat is. they are left in the audio stream. The dtmf-relay command
specifies how an H.323 or SIP gateway relays DTMF tones between telephony interfaces and
an IPnetwork. "Fhe complete command syntax is as follows:
dtmf-relay {[cisco-rtp] [h245-alphanumeric] [h245-signal] [rtp-nte
[digit-drop]] [sip-notify]}
Although all shown options are available when configuring a VoIP dial peer, only some of
them are applicable, depending on which signaling protocol isused. The options are as follows:
cisco-rtp(H.323 onl> ): Forwards DIMFtones using Real-Time Transport Protocol (RTP)
with a Cisco proprietary payload type
h245-alphanumeric (H.323 only): Forwards DIMF tones byusing the H.245
"alphanumeric" user input indication method: supports tones from 0 to 9. *. #. and from A
lo D: H.323 onl>
h245-signal (H.323 onl>): Fonvards DTMF tones by using the 11,245 "signal" user input
indication method: supports tones from 0 to 9. *. #. and from A to D
rtp-nte (H.323 andSIP): Fonvards DTMF tones byusing RTP with thenamed telephony
event (NTF) payload type
digit-drop (H.323 andSIP): Passes digits out-of-band anddrops in-band digits: only
available when the rtp-nte keyword is configured
sip-notify (SIP only }: Fonvards DTMF tones using SIPNotify messages: available only if
the VoIP dial peer is configured for SIP
Implemenling Cisco Voice Communicalions and QoS (CVOICE) v8 0
2010 Cisco Syslems. Inc
jgHVI
DTMF Relay Configuration Example
This subtopic provides a DTMF configuration example.
DTMF Relay Configuration Example
* Common methods: h245-alphanumerlc and rtp-nte
* Result: h245-alphanumeric selected as the higher priority choice
2001
2002
session target ipvi10.2.1.1
dtmf-relay h2*5-alphanumaric h2*5-signal rtp-nca
dial-peer voice i voip
dtmf-relay cisco-rtp h245-alphanumeric rtp-nte
This figure illustrates an example ofhow the DTMF relay methods are configured and
negotiated in H.323 and SIP.
H.323 is used for signaling and is the default protocol. During thecapabilities negotiation inthe
11.245 phase, the gateways exchange the supported DTMF relay methods. In this example, both
gateways support h245-alphanumeric and rtp-nte methods. Because h245-alphanumeric is
the higher priority choice, it isselected for all calls between the gateways. When adigit is
pressed on an endpoint telephone, itwill be signaled as an H.245 message, instead of
transmission in the RTP flow.
Configuring FAX/Modem Support
This topic describes how to configure support for fax andmodem transmissk
Cisco Fax Relay and Fax Pass-Through
Configuration
router(conf-voi-serv)
router (conf ig-dial -peer I tt
fax protocol {cisco ! none | system | pass-through
{g711ulaw j g711alaw}}
* Enables Cisco fax relay and pass-through globally (voice
service voip) or in dial-peer mode
System option exists indial-peer mode and refers to global
setting
Default for voice service voip: cisco
Default for dial-peer: system
The support for faxcan be defined usingthe following commands:
fax protocol: This command specifics if fax pass-through or Cisco laxrelay is negotiated,
and delines pass-through settings.
fax protocol t38: This command specifies if T.38 lax relay is negotiated and defines its
settings. This command ov envrites the fax protocol command, if issuedinthe same mode.
fax rate: This command can throttle down fax transmissionspeed.
fax-relay : This command enablesSuper Group3 (SG3) faxmachines to negotiate downto
G3 speeds.
Cisco Fax Relay and Fax Pass-Through
The fax protocol command is available globally (in voiceservice VoIP configuration mode),
andfor a specific dial peer(dial-peer configuration mode). Itenables eitherCisco fax relay or
fax pass-through, flicenabled option will benegotiated with theremote gateway before it can
be used. When fa\ pass-through is selected, the upspeed codecoptionsare: G.711 mu-law and
G.711 a-law,
Ihe dial peer settingtakesprecedence over theglobal setting, 'fhe global settingdefaults to
Ciscofax relay, whilethe dial peer settingdefaults to global setting.
*
T.38 Fax Relay Configuration
The fax protocol t38 command enables T.38 fax relay.
T.38 Fax Relay Configuration
router(conf-vol-serv)
router(config-dlal-peer)#
fax protocol t3fl [use [force]] [ls-redundancy value ths-
redundancy value]] [fallback {cisco | none | pass-through
{g711ulaw | g711alaw}}]
Enables T.38 fax relay globally (voice service voip) orin dial-peer
Overwrites fax protocol command (Cisco, none, pass-through)
Dial-peer setting takes precedence over global setting
Optionally, can use named signaling events (NSEs) conditionally or
unconditionally (force) to switch to T.38
Packet redundancy:
- Low-speed - from 0(default, no redundancy) to7copies
- High-speed - from 0(default, no redundancy) to3copies
Fallback options: Cisco proprietary fax relay, none, fax pass-through
The fax protocol t38 command is available globally (in voice-service VoIP configuration
mode) and for aspecific dial peer (dial-peer configuration mode). It overwrites the fax
protocol command, if issued in the same mode, because T.38 fax relay is mutually exclusive
with Cisco fax relay or pass-through. The dial peer setting takes precedence over the global
setting.
Theoptions arcthe following:
nse: Uses Named Signaling Events (NSKs) to switch to T.38 fax relay. The force keyword
uses NSEs unconditionally and is used for interoperability between H.323 or SIP. and
MGCP.
Is-redundancy: Specifies the number of redundant T.38 fax packets to be sent for the low-
speed V.21-based T.30 fax machine protocol. Range from 0to 7, default is 0.
hs-redundancy: Specifies the number ofredundant T.38 fax packets to be sent for high
speed V. 17. V.27. V.29. T.4. or T.6. Range from 0to 3, default is 0.
fallback: Afallback mode is used to transfer afax across aVoIP network ifT.38 fax relay-
could not besuccessfully negotiated at the time ofthefax transfer.
cisco: Asfallback option. Cisco proprietary fax relay.
pass-through: As fallback option, fax pass-through with cither G.711 mu-law or a-law
upspeed codec.
Fax Relay Speed Configuration
The fax rate command is used to establish the rate atwhich afax issent.
Fax Relay Speed Confiqural
rouCer(config-dial-peer)#
fax rate {2400 . 4800 7200 | 9600 | 12000 | 14400}
{disable voice} [bytes milliseconds]
' Throttles down fax transmission speed in dial-peer mode
Affects transmission length
disable: Disables fax relay transmission capability
voice: Highest possible transmission speed allowed by the
voice rate.
Can monopolize bandwidth
bytes: fax packetization rate, in milliseconds.
Range is 20 to 48. Default is 20
Fax relay transport is UDP. not RTP/UDP
RTP headercompression does not apply
J
The fax rate command can be configured for aspecific dial peer (dial-peer configuration
mode).
Ihe disable option disables fax relay transmission capability, fhe voice option selects the
highest possible transmission speed that is allowed by the codec rale.
The values for this command apply only to the fax transmission speed and do not affect the
quality ofthe fax itself. The higher transmission speed values (14.400 b/s) provide a faster
transmission speed but monopolize asignificantly large portion ofthe available bandwidth. The
lower transmission speed values (2400 b/s) prov ide aslower transmission speed and use a
relatively smaller portion ofthe available bandwidth.
1he tax eall is not compressed using the in r(p header-coinpression command because Simple
Packet Relay Transport (SPRT) o\cr User Datagram Protocol (UDP) is being used instead of
RTP. For example, a9600-b/s fax call takes approximatelv 24kb/s.
2-192 Implemenling Cisco Voice Communications andQoS (CVOICE] v8 0
*
Fax Relay SG3 Support Configuration
The fax-relay command allows for the suppression oftones from the fax machine side so that
SG3 fax machines can negotiate down toG3 speeds.
Fax Relay SG3 Support Configuration
router(conf-voi-oerv)
router(conflg-dial-pear)*
fax-relay {ans-disable | ecm disable | ag3-to-g3 [aystem] }
EnablesSuper Group 3 (SG3) faxmachinesto negotiate down to
G3 speeds (from upto 33.6 kb/s to up to 14.4 kb/s)
Applicable to Ciscoproprietary and T.38 faxrelay
ans-disable: Disables ANStones fromoriginatingin SG3 fax
machines so that these machines can operate at G3 speeds
ecm disable: Suppresses Error Correction Mode
SG3-to-G3: AllowsSG3 machines to negotiate down to G3 speeds
System option exists indial-peer mode and refers toglobal setting
Default Not enabledmodem upspeed can occur when ANStones
are detected, fax-relay ECM is enabled, and SG3-to-SG3fax relay
communication is not supported and probably will fail
The fax-relay command is also used to disable fax-relay Error Correction Mode (ECM). The
command isconfigured globally (in voice-service VoIP configuration mode) orindial-peer
configuration mode. The dial-peer mode has the system keyword torefer to the global setting.
The ans-disable option suppresses answer (ANS) tones from originating SG3 fax machines so
that these machines canoperate at G3speeds using fax relay.
The ecm disable option disables fax-relay ECM.
The sg3-to-g3 option allows SG3 machines tonegotiate down toG3 speeds using fax relay.
If the fax-relay command is notconfigured, modem upspeed can occur when ANS tones arc
detected, fax-relay ECM isenabled, and SG3-to-SG3 fax relay communication isnot
supported. Thefax communications will probably fail.
Fax Support Configuration Example
This subtopic provides a fax support configuration example.
IX bu
Result: Cisco fax relay, fax rate 4800 b/s, SG3 support
fax rate 4S00
fax-relay ecn disable
fax-relay sg3-to-g3
fan-relay ans-disable
fax rate 4800
IP WAN
R2
102 11 :-,-.... 2001
*ss^n=5O/0/i
fax-relay ecm disable
fax-relay sg3-tc.-g3
fax-relay ans-disable
fax rate 7200
fax protocol t3B ls-redundancy 2
hs-redundaney 2 fallback cisco
The figure shows two gateways with dial peers configured for fax support. R2 isconfigured for
T.38 fax relay with a fallback option toCisco fax relay. Rl uses the default fax protocol setting,
which is Cisco fax relay . Cisco fax relay is negotiated between thegateways when a fax
transmission occurs, R2 throttles down to 7200 b/s, so the lowest common value is 4X00 b/s
(faxrate of Rl). Both gateways areconfigured tosupport SG3 fax machines so thatthey will
negotiate the transmission speed down to G3.
Configuring Modem Support
This topic explains how toconfigure modem support.
Modem Pass-Through
router(config-dial-peer)t
modem passthrough {system | nse [payload-type number
codec {g711ulaw | g711alaw} [redundancy]
_Z3
Enables modem pass-through globally (voice-service VoIP
configuration mode) or in dial-peer mode
Default: disabled
Dial-peer setting takes precedence overglobal setting
system: refers to global setting
nse: NSE used tosignal switchover to modem pass-through
redundancy: single repetition of packetsto protect against
packet loss
Modem pass-through and relay arc configured using three commands as follows:
modem passthrough: This command enables modem pass-lhrough.
modem relay: This command enables modem pass-through or relay, depending on the
negotiation results. It removes the modem passthrough command, ifconfigured in the
same mode.
modem relay gateway-xid: This command configures additional modem relay parameters,
such as compression.
Modem Pass-Through
Modem pass-through can be configured globally (in voice-service VoIP configuration mode) or
in dial-peer configuration mode using the modem pass-through command. The system option
isavailable in the dial-peer mode and references the global setting. The nse option defines that
NSts are usedto communicate codecswitchover between gateways, with theoptional
specification ofthe payload type. Ifthe payload type is configured explicitly, it must be set to
the same value on both the originating and tenninating gateway. The codec option defines the
upspeed codec. The redundancy option enables asingle repetition ofpackets to improve
reliability by protecting againstpacket loss.
Modem Relay
Configuring Modem Suj
Modem Relay
routerlconf-voi-aerv|
router(con fig-dial-peer)#
modem relay {nse [payload-type number] codec {g711alaw
g711ulaw} [redundancy] | system} gw-controlled
Enables Cisco proprietary modern relay or pass-through, depending on
the negotiation
- Configured globally (voice-service VoIP configuration mode) or indial-
peer mode
- Dial-peer settingtakes precedence over global setting
- nse NSEused to signal switchover to modemrelay
codec For fallback to pass-through
redundancy Single repetition of packets
system Refers to global setting
* gw-controlled Selects gateway-controlled method
1he modem rela> command enables modem pass-through orrelay, depending onthe
negotiation results. It removes the modem passthrough command, if configured in the same
mode. Modem relay can beconfigured globally (in voice-service Voll' configuration mode) or
indial-peer configuration mode. The system option isavailable inthe dial-peer mode and
references the global setting.
fhe nse option defines that NSFs are used lo communicate codec switchover between
gateways, with theoptional specification ofthe NSF payload type. Range varies by platform,
and is typicall} from 98 to 117. If thepayload type is configured explicitly, it must beset tothe
same value onboth theoriginating and terminating gateway. Thecodecoption defines the
upspeed codec, which is used when pass-through is negotiated and relay is not. The
redundancy option enables a single repetition of packets when pass-lhrough is negotiated and
relay isnot. The gw-controlled option selects the gateway-configured method for establishing
modem relay parameters.
2010 Cisco Systems, inc.
Modem Relay Compression
router(conig-dial-peor)#
modem relay gateway-xid [compress {backward | both
forward | no}] [dictionary value] [string-length value]}
compress: directioninwhich data is compressed
- Default: both
dictionary value and string-length value: characteristics ofthe
compression algorithm
- Default: 1024 and 32, respectively
Default: enabled when modem relay NSE is enabled
Modem Relay Compression
The modem relaygateway-xid command configures in-band negotiation of compression
parameters between two VoIP gateways. This setting can beconfigured globally (in voice-
service VoIP configuration mode) or indial-peer configuration mode, 'fhe dial-peer setting has
higher precedence than the global setting. The command isenabled when the modem relay
command is configured.
Thecompress option specifies thedirection inwhich data flow iscompressed. Fornormal
operations, compression should beenabled inboth directions. This isthe default setting.
Forward compression isused onthe originating gateway toreduce the amount ofdata that is
sent towards theterminating gateway. Backward compression is the ability ofthe terminating
gateway to correctly interpret the compressed data that isreceived from the originating
gateway. Forward compression ononegateway must bematched bybackward compression on
the peer gateway, fhebackward parameter enables compression only inthe backward
direction. The forward parameter enables compression only inthe forward direction. Theno
parameter disables compression in both directions.
Thedictionary and string-length options define the V.42 bisparameters thatspecify the
compression algorithm characteristics. Range isfrom 512to2048, and 16 to32, respectively.
Defaults are 1024 and32.respectively. Modems may support values higher than these ranges.
Avalue acceptable tobothsides is negotiated during modem call setup.
Modem Pass-Through and Modem Relay Interaction
Ihis subtopic explains the interaction between modem pass-lhrough and modem relay.
fem Pass-Through and Modem
Relay Interaction
Bothgateways must be configured for Cisco modem relay.
Ifone is configured for pass-through, pass-through is used.
* Cisco modem relay is not negotiated
Parameters are configured on the gateway (gateway-controlled).
NSEs indicate switchover:
From voice to modem pass-through (voiceband data)
From modem pass-through to modem relay
Upon detecting 2100-Hz tone, the terminating gateway sends:
NSE 192 and switches over to modem pass-through
NSE 199 to indicate modem relay:
Ifrecognized by the originating gateway, use modem relay
Ifnot recognized, use modem pass-through
Ciscomodem relay is a nonnegotiated. bearer-sw itched modefor modem transport that does
not involve eall agent-assisted negotiation duringthe call setup. Instead, the negotiation
parameters are configured directly on the gateway. Thesegateway-controlled negotiation
parameters use \SFs to indicate the switchover from voice, to voiceband data (VBD). to
modem relay.
Upon detecting a 2100-11/ tone, the tenninatinggateway sendsan NSF 192 to theoriginating
gateway and switches over to modem pass-through. The terminating gateway also sends an
NSF 199 to indicate modem relay. If this event is recognized by the originating gateway, the
call occursas modem relay. If the event is not recognized, the call occursas modem pass-
through.
Incaseof MGCP signaling, because modem relay has configured locally on the gateways, it
removes the signaling dependency fromthe call agent and allows modemrelay support
independent of call control. The gateway-controlled modem relayparameters are enabled b>
default when Ciscomodem relay is configured, and whenCiscomodem relay is configured,
gateway exchange identification(XID) parameter negotiation is always enabled. Gateway XID
parameters arc negotiated using the SPRT protocol.
2-198 Implementing Cisco Voice Communications and QoS (CVOICE] v8.0 2010 Cisco Systems. Inc.
Modem Support Configuration Example
This subtopic provides amodem support configuration example.
Modem Support Configuration Example
Result: modem pass-through, G.711 mu-law, one-way
redundancy
session target ipv4sl0.2.1.1
modem passthrough nse codec glllulaw radundancy
destinntion-pattern 100.
session target lpv4:10.1.1.1
modem relay noe codec g711ulaw gv-controllad
The figure shows two gateways that are configured tosupport modem transmission over an IP
network. Rl isconfigured for pass-through while R2 isconfigured for modem relay and pass-
through. Both gateways agree onmodem pass-through with upspeed codec set toG.711 mu-
law. Redundant packets will besent only inonedirectionfrom Rl to R2.
Configuring Codecs
Ihis topic describes howto configure codec negotiation in the VoIP environment.
Configuring Codec Li:
router(config)#
voice class codec class tag
Creates a codec voice class
router(config-claes)#
codec preference value codec-type [mode frame-size][bytes
payload-size]
Configuresthe codec voiceclass with codecs and their preferences
Mode and frame size apply to iLBC"
20' 20-ms frames for 15 2 kb/s bit rate (default)
30 30-ms frames for 13 33 kb/s bit rate
Payload size voice payload of each frame
Values depend on the codec type
* Additional options exist for GSMAMR-NB codec
Cisco voice gateway s offertheoption todefine a listof codecs to be used for negotiation of
VoIP capabilities.
A codec list is configured as a codec \oiee class using the \ oice class command and identified
using a class-tag.
Thecodecmice class allows the configuration ol'a prioritized list of codecsand their
parameters. Ihe preference valuerepresents the priority ol'a givencodectype.
fhe modeand frame-si/e parameters apply to Internet Fow Hilratc Codec (iFBC) and signify
the following:
20: 20-ms frames for 15.2 klv's bit rate (default)
30: 30-ms frames for 13.33 kh's bit rate
The payload-size parameter defines the voice payload of each frame. 1he available values
dependon the selected codec type.
2-200 Implementing Cisco Voice Communications and QoS (CVOICEI v3 0 2010 Cisco Systems, Inc.
Codec-Related Dial Peer Configuration
This subtopic describes how to apply codec-related settings to dial peers.
Codec-Related Dial Peer Configuration
router(config-dlal-peer)*
/oice-closs codec claaatag
Assignscodec voice class to dialpeer (multiple codec option)
router(conflg-dial-peer)W
codec {codec [bytes payload-size]
bytes]
transparent} [fixed-
Defines an individual codec on a dial peer
payload size: voice payloadof each frame
- Values depend on the codec type
transparent:enables codec capabilities to be passed transparently
between endpoints in a Cisco Unified Border Element
fixed-bytes: codec byte size is fixed and nonnegotiable
Default: g729r8, 20-byte payload
Thecodec settings areapplied to VoIP dial peersusingeitherof twoways:
Thevoice-class codec command applies a list of codecs that areconfigured withthevoice
class codeccommand. Thisoption enables multiple codec types for thegiven dial peer.
The codec command specifies a single codec tobeused bythegiven dial peer. The default
is g729r8. 20-byte payload. The options for the single codec include the following:
payload-size: voice payload ofeach frame; available values depend on the codec
type
transparent: enables codec capabilities lobepassed transparently between
endpoints ina Cisco Unified Border Element
il\ed-bytes: codec byte size is fixed and nonnegotiable
i 2010 Cisco Systems. Inc
Codec Configuration Example
This subtopic pro\idesa codec configuration example.
Codec Configuration Ex*
codec prefe
codec prefe
codec prefe
ii-pe
IP WAN
100
1 g711alaw
2 g711ulBw bytes 80
3 g723ar53
4 g723ar63 bytes 144
5 g723rS3
6 g723c63 bytes 120
7 g726rl6
S g726r24
9 g726r32 bytes 80
10 g?28
11 g729brS
12 g729r3
rget ipv4:10.2.1.1
session target ipvti10.1.1.1
Result: g729r8, 20 bytes
Inthe figure, twogateways negotiate calls using H.323. When Rl signals a call, it offers a large
set of supported codecs, configured using the \ nice-class codec command. When R2 receives
thecall setup request, it matches the inbound dial peer. Inibisexample, theinbound dial peer is
dial peer 4. which supports only the default codec G.729rX with 20-byte payload. Il'dialpeer4
didnot exist on R2. R2 would match thedefault dial peer(dial peer0). Because thedefault dial
peer supports all codecs. R2wouldselect the first codec intheofferedproposal (G.711 a-law).
2-202 Implementing Cisco Voice Communicalions and QoS (CVOICE) u8.0
j*Q0
Limiting Concurrent Calls
This topic describes how tolimit the number ofconcurrent calls.
Limiting Concurrent Calls
router(config-dial-paer}#
max-conn number
Specifies the maximumnumber of incoming or outgoing
connections for a particular dial peer
Typically used todefinethe numberofconnections used
simultaneouslyto send or receive fax-mail, for off-ramp
store-and-forward fax functions
- Can be applied to these dial peer types: POTS, VoIP.
Multimedia Mail over IP(MMolP), or Voice over Frame Relay
(VoFR)
Number range: 1 to 2,147,483,647
Default: no limit
Thetotal number of eitherincoming or outgoing connections canbe limited on theper-dial peer
basis. This feature is typically used todefine thenumber of connections thatareused
simultaneously losend or receive fax-mail, for off-ramp store-and-forward fax functions. The
limit is configured using the max-conn command inthedial-peer configuration mode. By-
default, no limit is imposed.
Summary
Ihis topicsummarizes the keypointsthat werediscussed in this lesson.
Summary
VoIP dial peers relate to voice call legs established over the
IP WAN.
DTMF relay options (cisco-rtp, h245-alphanumeric, h245-
signal, rtp-nte, and sip-notify) are configured in the dial
peer, and are subject to negotiation.
T.38 faxrelay can fall backto Ciscoproprietary faxrelayor
fax pass-through, while Ciscoproprietary modemrelaymay
fall back to modem pass-through, if not supported bythe peer
gateway.
Dial peers can be configured with a prioritized codec list or a
single codec
The number of calls can be limited in the dial peer
2-204 Implementing Cisco Voice Communications and QoS (CVOICE) vB.i
Module Summary
Ihis topic summarizes the key points that were discussed in this module.
Module Summary
VoIP transmission requires the sampling, encoding,
packetization, VoIP transmission in RTPor its variant,
decoding, and modulation.
Gateways using peer-to-peer signaling protocols (H.323, SIP)
build the dial plan using the dial peers.
SIP is an RFC-based signaling protocol with open
architecture that allows flexibility and extensibility.
MGCP gateways forward calls by receiving instructions from
call agent and responding to its requests.
Audiotransmission quality depends on factors such as delay,
jitter, packet loss, and available bandwidth.
VoIP dial peers can be configuredto support fax/modem
pass-through, relay, and DTMF relay.
This module explains how VoIP technology differs from circuit-based telephony. It describes
thesteps that thegateways perform toconvert the voice wavelength toa stream of VoIP
packets, transmit the information using Real-Time Transport Protocol (RTP) orits variants, and
recreate thevoice wavelength. Special attention is given tothe characteristics andproper
implementation ofVoIP gateway signaling protocols: H.323, Session Initiation Protocol (SIP),
and Media Gateway Control Protocol (MGCP). The module defines theQoS-relatcd parameters
that audio transmission demands from thetransport IPnetwork andexplains methods for
transporting special data types, such as fax, modem, and dual lone multifrequency (DIMF)
tones. Finally, it covers the configuration of VoIP dial peers, including advanced features such
as fax/modem support and DTMF relay.
2-206 Implemenling Cisco Voice Communications and QoS (CVOICE) vS.O 2010 Cisco Systems, Inc.
Module Self-Check
Use the questions here to review what you learned in this module, 'fhe correct answers and
solutions arefound in the Module Self-Check Answer Key.
Q1) Put the major stages of VoIP processing in the correct order. (Source: Examining Vol P
Call Leg Characteristics and VoIP Media Transmission)
modulation
____ decapsulation
sampling
^transport overIPnetwork
_____ encoding
decoding
quantization
codec compression
encapsulation
Q2) Pair the processes and corresponding parameters. (Source: Examining VoIP Call Leg
Characteristics and VoIP Media Transmission)
A) compression
B) encoding
C) sampling
D) transport over IP network
1. PPP overhead
2. a-law
3. iLBC
4. 50 Hz
Q3) What isthe uncompressed Layer 3+ bandwidth ofaG.729 call with a packetization rate
of25 p/s? (Source: Examining VoIP Call Leg Characteristics and VoIP Media
"I ransmission)
A) 20 kb/s
B) 24 kb/s
C) 18 kb/s
D) 16 kb/s
Q4) What isa function ofRTP? (Source: Examining VoIP Call Leg Characteristics and
VoIP Media Transmission)
A) call multiplexing
B) encryption
C) payload identification
D) replay protection
2010 Cisco Systems. Inc VoIP Call Legs 2-207
Q5) VAD kactive by default in G.729A calls. (Source: Examining VoIP Call Leg
Characteristics and VoIP Media Transmission)
A) true
B) false
Q6) Which two tasks are performed by the RAS signaling function of H.225?
(Choose two.) (Source: Explaining 11.323 Signaling Protocol)
A) conducts bandwidth changes
B) transports audio messages between endpoints
C) conducts disengage procedures between endpoints and agatekeeper
D) allows endpoints tocreate connections between call agents
F] defines call setup procedures that are based onISDN call setup
Q7) Match the 11.245 control function with its description. (Source: Explaining 11.323
Signaling Protocol)
A) logical channel signaling
B) capabilities exchange
C) master/sla\ e detennination
D) mode request
.... I- opens and closes the channel that carries the media stream
2. asks for a change in capability ofthe media stream
3. negotiates audio, video, andcodec capability between theendpoints
4. resolves conflicts during the call
Q8) What isthe lunctionofan H.323 gatcwaj? (Source: Explaining 11.323 Signaling
Protocol)
A) comert an alias address to an IP address
B) respond to bandw idth requests and modifications
C) transmit and receive G.711 I'CM-encoded voice
D) perform translation between audio, video, and data formats
E) recei\e andprocess multiple streams of multimedia input
09) With 11.323 call establishment, which channel is used byendpoints to communicate
with the gatekeeper when establishing an H.323 call? (Source: Explaining H.323
Signaling Protocol)
A) B channel
B) RAS channel
C) forward channel
D) in-band control channel
Implementing Cisco Voice Communications and QoS(CVOICE) v80 2010CiscoSystems, Inc
Q10) Put the steps that are involved in 11.323 basic call setup without agatekeeper in the
correct sequence. (Source: Explaining H.323 Signaling Protocol)
A) Call setup procedures that are based on Q.931 create a call-signaling channel
between the endpoints.
B) The H.245 control function negotiates capabilities and exchanges logical
channel descriptions.
C) The gateway determines the IP address ofthe destination gateway internally.
D) The logical channel descriptions open RTP sessions.
E) The endpoints open another channel for theH.245 control function.
F) The originating gateway initiates an H.225 session with the destination
gateway onregistered TCPport 1720.
G) The endpoints exchange multimedia over the RTPsessions.
1. Step 1
_____ 2. Step 2
3. Step 3
4. Step 4
5. Step 5
6. Step 6
7. Step 7
Q11) How does the abbreviated call setup procedure in H.323 version 2 provide fast start?
(Source: Explaining H.323 Signaling Protocol)
A) Thegateway knows a DNS-rcsolvable domain name for thedestination.
B) Endpoints use a separate channel for H.245 control functions to speed up
signaling.
C) Capability exchange andlogical channel assignments arecompleted inone
round trip.
D) Endpoints andgateways usethesame call control model sothat no translation
is required.
Q12) In Gatekeeper-Routed Call Signaling, what is the role of the gatekeeper? (Source:
Explaining H.323 Signaling Protocol)
A) establishes an H.245control channel with both endpoints
B) performs call setup, call function, and call control functions
C) represents the other endpoint for call signaling
D) passes on the request fromthe originating endpointto the terminating endpoint
Q13) Which configuration is required to activate anH.323 gateway on a Cisco router?
(Source: Explaining H.323 Signaling Protocol)
A) gateway command in global configuration mode
B) setting the gateway source IP address
C) binding the gateway functionalityto an interface
D) none (11.323 gateway is enabled by default)
Q14) TCPis the default transport protocol on a Ciscogateway. (Source: Explaining H.323
Signaling Protocol)
A) true
B) false
12010 Cisco Systems, Inc. VoIP Call Legs 2-209
QI5) fhe _____ command is used toverify that the H.323 gateway is operational and
displays the status of the gatewav. (Source: Explaining H.323 Signaling Protocol)
QI6) Which SIP clement selects ihe media tvpe and parameters? (Source: Explaining SIP
Signaling Protocol)
A) user location seniccs
B) user capabilities services
C) user availabilitv services
D) call setup services
E) call handling services
Q17) What are three advantages of SIP? (Choose three.) (Source: Explaining SIP Signaling
Protocol)
A) dial-plan configuration directly on the gateway
B) support of third-party end devices
C) interoperability with H.323
I)) translations defined per gatewav
1.) ITU standard
I ) minimum bandwidth usage
Q18) Which four are SIP servers? (Choose four.) (Source: Explaining SIP Signaling
Protocol)
A) registrar
B) gatewav
C) redirect
D) location
K) proxv
1) database
G) relocatio
Q19) Which message and contents does the IPphonesendto the SIPproxyserver whena
user initiates a SIP call? (Source: ExplainingSIP Signaling Protocol)
A) INVITE, the E.164 number ofthe called party
B) INVITE, the E.164number ofthe SIP gateway
C) ACK. the F. 164 number ofthe PS'I'Nphone
D) OK. the H.164 number ofthe PSTNphone
Q20) Which tvpeof SIPaddress is represented by the entry
"sip: 19193631234 ^gateway.com:user=phone"? (Source: Explaining SIPSignaling
Protocol)
A) a fully qualified domain name
B) an E.164 address
C) a mixed address
D) a ERE address
Q2I) What is one disadvantage ofthe direct call setupmethod? (Source: Explaining SIP
Signaling Protocol)
A) It relies on cached information, which may be out of date,
B) It uses more bandwidth because it requires more messaging.
C) It must learn the coordinates ofthe destination 1IA.
D) It needs the assistance of a network server.
2-210 Implemenling Cisco Voice Communications and QoS (CVOICE) v8 0 2010 Cisco Systems. Inc.
JMV
Q22) What is the characteristic ofSIP early offer? (Source: Explaining SIP Signaling
Protocol)
A) The media channel issetup before the call isanswered.
B) The recommended use iswith SIP trunks to allow the ITSP toprovide their
capabilities first.
C) The INVITE message carries SDP.
D) The signaling exchange uses fewer messages than indelayed offer.
Q23) What isthe transport ofSR'fP flows when both signaling and media are secured (SIPS
andSRTP)? (Source: Explaining SIPSignaling Protocol)
A) IP/ESP
B) IP/TCP/TLS
C) IP/UDP
D) 1P/TLS
Q24) What does the session target sip-server dial-peer command do? (Source: Explaining
SIP Signaling Protocol)
A) It tells the routerto use DNSto resolvesip-server.
B) It tells the routerto use the server that is identified inthe SIP UA
configuration.
C) It tells the router to useSIPas the sessionprotocol.
D) fhis is invalid syntax, and an error will be generated.
Q25) Which debugcommand would youuse totrace call setups, connections, and
disconnections? (Source: Explaining SIPSignaling Protocol)
A) debug voip ccapi inout
B) debug ccsip calls
C) debug ccsip states
D) debug ccsip messages
E) debug ccsip events
Q26) Which control model is used with MGCP? (Source: Explaining MGCP Signaling
Protocol)
A) gateway-oriented
B) distributed
C) peer-to-peer
D) stimulus
Q27) Which protocol does MGCP use to describe thetypeof initiated session? (Source:
Explaining MGCP Signaling Protocol)
A) SIP
B) Cisco Discovery Protocol
C) SDP
D) MGC
Q28) Because thegateway configuration is mostly done onthe , far fewer CiscoIOS
router commands are necessaryto bring up the gateway than in any other gateway type.
(Source: Explaining MGCP Signaling Protocol)
)2010 CiscoSystems, Inc. VoIPCall Legs 2-211
029) Which MGCP component represents the place ofinterconnection between the packet
network and the traditional telephone network? (Source: Explaining MGCP Signaling
Protocol)
A) endpoint
B) natewav
C) gatekeeper
D) call agent
Q30) What is the function of an MGCP gateway? (Source: Explaining MGCP Signaling
Protocol)
A) routes calls to locally attached endpoints
B) manages thetranslation of audio between theSCN andpacket-switched
network
C) controls the operation ofthe endpoints andthe call agent
D) allows authenticated traffic into the network
031) Which two messages are issued bv a gateway? (Choose Iwo.) (Source: Explaining
MGCP Signaling Protocol)
A) AuditConnection
B) NotifiealionRequest
C) CreateConnection
IJ) DeleteConnection
I ) RestartlnProgress
032) Match the MGCP control message with its function. (Source: Explaining MGCP
Signaling Protocol)
A) AuditConnection
B) NotificationRequest
C) Modifv Connection
D) AuditF.ndpoint
E) RestartlnProgress
1. requests the status of an endpoint
2. instructs the gateway on which action to take on the occurrence of an event
3. requests the status of a connection
4. notifies the call agent that the gateway and its endpoints are removed from
sen ice
-^ instructs the gateway lo update its connection parameters for a prevtouslv
established connection
033) In an MGCPenvironment. the call agent sends a to each gateway as the first
event in the call setup process. (Source: Explaining MGCPSignaling Protocol)
034) Which configurationcommandenables MGCPon UDP port 5000? (Source:
Explaining MGCP Signaling Protocol)
A) mgcp 5000 global configuration command
B) mgcp udp 5000 global configuration command
C) mgcp 5000 interface configuration subcommand
[)) mgcp gatewav 5000 global configuration subcommand
Q35) How do you configure arouter to use MGCP on adigital port? (Source: Explaining
MGCP Signaling Protocol)
A) Add the application mgcpapp subcommand tothe dial peer.
B) Add theservice mgcpsubcommand to thedial peer.
C) Add the application mgcpapp parameter tothe dsO-group controller
command.
D) Add the service mgcp parameter tothe dsO-group controller command.
Q36) Which command displays alist ofthe voice ports that are configured for MGCP?
(Source: Explaining MGCP Signaling Protocol)
A) show mgcp gateway
B) show endpoints
C) show mgcp endpoint
D) show mgcp
037) Which command displays a count ofthesuccessful and unsuccessful control
commands? (Source: Explaining MGCP SignalingProtocol)
A) show mgcp calls
B) show mgcp statistics
C) show mgcp
D) debug mgcp statistics
Q38) Jitter isdefined asa variation inthe ofreceived packets. (Source: Describing
Requirements for VoIP Call Legs)
039) What aretwogeneral transport requirements in VoIP networks? (Choose two.)
(Source: Describing Requirements for VoIPCall Legs)
A) maximum 300-ms two-way delay
B) maximum 1 percent packet loss
C) maximum 180-msone-way delay
D) maximum 30-ms one-way jitter
E) maximum 50-ms two-way jitter
F) 17-to 80-kb/scall bandwidth including I.aycr 3+overhead
Q40) fhe three conceptual methods of carrying virtual real-time fax machine-to-fax machine
communications acrosspacket networksare fax _, and . (Source:
Describing Requirements for VoIP Call Legs)
Q41) Fax is thesimplest technique for sending a faxover IPnetworks. (Source:
DescribingRequirements for VoIP Call Legs)
Q42) Control of faxpass-through is achieved through _ that are sent in the RTP stream,
(Source: Describing Requirements for VoIP Call Legs)
Q43) The tone that is generated ona touch-tone phone when thekeypad digits arepressed is
called . (Source: Describing Requirements for VoIP Call Legs)
Q44) What is achieved byconfiguring loopback addresses in thesession target command?
(Source: Configuring VoIP Call Legs)
A) The calls fail if ihe interface goes down.
B) The interface will never shut down.
C) The calls will use an alternate path if the interface shuts down.
D) The calls will never fail as long as the router is operating.
i 2010Cisco Systems, Inc. VoIP Call Legs 2-213
Q45) When the router finds a matching outbound VoIP dial peer, which command
determines where to forward the call? (Source: Configuring VoIP Call Legs)
A) destination-pattern
B) port
C) session target
D) dialplan number
046) What happens when gatewavs fail tonegotiate acommon DTMF relay method?
(Source: Configuring VoIP Call Legs)
A) DIMF tones are dropped.
B) DTMF tones are left in-band.
C) DTMF tones are let) out-of-band.
D) DIMFtones arecarried asymmetrically, using the method that is preferred by-
each gatewav.
Q47) Which statement about bandwidth requirements for fax relay is true? (Source:
Configuring VoIP Call Legs)
A) Faxrelav transmission bandwidth can be savedby usingRTP.
B) Ihe l.aver3* overhead (IP/UDP/RTP) without compression is 40 bytes.
C) The transmission bandwidth musl bethrottledto avoidmonopolization of
network resources.
D) Random dropsof faxpackets slowdownthe transmission speed,
Q48) Which codecwill be negotiated when theoriginating gateway offers(i.729 (preference
I) and G.7! 1(preference 2). and the tenninating gateway offers vice versa (G.711 with
preference 1and G.729withpreference 2)? (Source: Configuring VoIPCall Legs)
Q49) What is the tvpical reason to limitthe numberof concurrent calls?(Source:
Configuring VoIP Call Legs)
A) saving gatewav processingpower
B) controlling the number of calls that are sent through the WAN
C) limiting the number of off-ramp store-and-forward fax transmissions
D) enforcing the enterprise policies
2-214 Implemenling CiscoVoice Communications and QoS (CVOICE) v80 2010 CiscoSystems, Inc
QD
1 sampling
2 quantization
3 encoding
4 codec compression
5 encapsulation
6 decoding
7 transport over IP network
8. decapsulation
9. modulation
Q2) J-A
2-B
4-C
l-D
Q.U D
Q4) C
05) B
06) A.C
Q7)
l-A
2-D
.1-B
4-C
Q8) D
09) B
QIO) 1-F
2-C
3-A
4-E
5-B
6-D
7-G
QU) C
Q12) C
Q13) D
014) A
015) show gateway
Q16)
B
QI7) A, B.D
QI8) A. CD, F
QI9) A
Q20) B
Q2I) A
Q22) C
Q2.U C
Q24) B
Q25) B
>2010 Cisco Systems, Inc. VoIPCall Legs 2-215
02(il
D
02') C
02S| Cisco Unified Communications Maiutiiei
02'>) A
Q.-0) H
031)
B. 1-.
032) 1-1)
2-B
3-A
4-E
5-C
033) Notification Request (RON'!")
0--4) A
035) I)
QMi) C
Q3 "i li
O'S) Llcl;i>
039) b. n
040) pass-through, relav storc-and-lorârd
041 1 pass-tli rousill
042) Named Sijm;ihn_i! l-venis (NSEs 1
043) du.il lone iiiultifiequeiicv (DI'MI-)
044) I")
045) C
046) B
0471 C
Q4S) CI 1
049) c
2-216 Implementing Cisco Voice Communicalions and QoS (CVOICE| v8 0 2010 Cisco Systems, Inc
Table of Contents
Volume 2
Cisco UnifiedCommunications ManagerExpress Endpoints Implementation 3-1
Overview 3-1
Module Objectives 3-1
Introducing Cisco Unified Communications Manager Express hi
Objectives 3-3
Introducing Cisco Unified Communications Manager Express 3-4
CiscoUnified Communications ManagerExpress Positioning 3-5
CiscoUnified Communications ManagerExpress Deployment Models 3-6
Cisco Unified Communications ManagerExpress Key Features and Benefits 3-8
Phone Features 3-9
SystemFeatures 3-11
Trunk Features 3'13
Voice-Mail Features 3-14
CiscoUnified Communications ManagerExpress SupportedPlatforms 3-15
Cisco Integrated Services Routers Scalability 3-16
Cisco Integrated Services Routers Generation 2 Scalability 3-17
Memory Requirements 3-18
Cisco Integrated Services Routers Licensing and Software 3-19
CiscoIntegrated Services Routers Generation 2 Licensing Model 3-20
Cisco Unified Communications Manager Express Operation 3-21
Calls Through IPNetwork 3-22
Summary 3-23
Examinino Cisco Unified Communications Manager Express
Endpoint Requirements , , 3-25
Objectives 3-25
Overview of Cisco Unified Communications Manager Express Endpoints 3-27
Endpoint Signaling Protocols 3-28
Endpoint Capabilities 3-29
Basic Cisco Unified IP Phone Models 3-30
Midrange Cisco Unified IPPhones 3-31
Upper-End CiscoUnified IPPhones 3-32
Video-Enabled Cisco Unified IP Phones 3-33
Conference Stations 3-34
Identifying Cisco Unified Communications ManagerExpress Endpoint Requirements 3-35
Phone Startup Process 3-37
Power over Ethernet 3-43
Two PoE Technologies 3-44
Cisco Prestandard Device Detection 3-46
IEEE 802.3af Device Detection 3-47
CiscoCatalystSwitch: Configuring PoE 3-48
Cisco Catalyst Switch: Show Inline Power Status 3-49
VLAN Infrastructure 3-50
VoiceVLAN Support 3-52
SingleVLAN Access Port 3-53
Multi-VLANAccess Port 3-54
Trunk Port 3-56
Ethernet Frame Types Generated by Cisco IP Phones 3-57
Blocking PCVLAN Access at IPPhones 3-58
Limiting VLANs on Trunk Ports at the Switch 3-59
Configuring VoiceVLAN inAccess Ports UsingCisco IOS Software 3-60
ConfiguringTrunk Ports Using Cisco IOS Software 3-62
Verifying Voice VLAN Configuration Using Cisco IOS Software 3-64
IPAddressing and DHCP 3-65
DHCP Parameters 3-66
Router Configuration with 802.1Q Trunk 3-67
Router Configuration with Cisco EtherSwitch Network Module 3-68
DHCP Relay Configuration 3-69
Network Time Protocol 3-70
NTP Configuration Example 3-71
Endpoint Firmware and Configuration 3-72
Downloading Firmware 3-73
Firmware Images 3-74
Setting up Cisco Unified Communications Manager Express in SCCP Environment 3-75
Configuring Source IP Address and Firmware Association 3-76
Enabling SCCP Endpoints 3-78
Locale Parameters 3-79
Date and Time Parameters 3-80
Parameter Tuning 3-81
Generating Configuration Files for SCCP Endpoints 3-82
Cisco Unified Communications Manager Express SCCP Environment Example 3-84
Setting Up Cisco Unified Communications Manager Express in a SIP Environment 3-85
Configuring Cisco Unified Communications Manager Express for SIP 3-87
Configuring Source IP Address and Associating Firmware 3-88
Enabling SIP Endpoints 3-89
Locale Parameters 3-90
Date and Time Parameters 3-91
NTP and DST Parameters 3-92
Generating Configuration Files for SIP Endpoints 3-93
Cisco Unified Communications Manager Express SIP Environment Example 3-94
Summary 3-95
Configuring Cisco Unified Communications Manager Express Endpoints 3-97
Objectives 3-97
Directory Numbers and Phones in Cisco Unified Communications Manager Express 3-99
Directory Number Types 3-100
Single and Dual-Line Directory Numbers 3-101
Octo-Line Directory Number 3-102
Nonexclusive Shared-Line Directory Number 3-103
Exclusive Shared-Line Directory Number 3-104
Multiple Directory Numbers with One Telephone Number 3-105
Multiple-Number Directory Number 3-106
Overlaid Directory Number 3-107
Creating Directory Numbers for SCCP Phones 3-108
Single-Line Ephone-dn Configuration 3-110
Dual-Line Ephone-dn Configuration 3-111
Octo-Line Ephone-dn Configuration 3-112
Dual-Number Ephone-dn Configuration 3-113
Configuring SCCP Phone-Type Templates 3-114
Configuring SCCP Phone-Type Templates 3-115
Ephone Template for Conference Station 7937G Configuration Example 3-117
Creating SCCP Phones 3-118
Configuring SCCP Ephone Type 3-119
Configunng SCCP Ephone Buttons 3-120
Configuring Ephone Preferred Codec 3-121
Basic Ephone Configuration Example 3-122
Multiple Ephone Configuration Example 3-123
Multiple Directory Numbers Configuration Example 3-124
Shared Directory Number Configuration Example 3-125
Controlling Automatic Registration 3-126
Partially Automated Endpoint Deployment 3-127
Partially Automated Deployment Example 3-128
Creating Directory Numbers for SIP Phones 3-129
Voice Register Directory Number Configuration Example 3-130
Creating SIP Phones 3-131
Configuring SIP Phones 3-132
Tuning SIP Phones 3-133
Implementing Cisco Voice Communications and QoS (CVOICE) wB.O 2010 Cisco Systems. Inc
Shared Directory Number Configuration Example 3"134
Configuring Cisco IP Communicator Support 3-135
Configuring Cisco IP Communicator 3-136
Managing Cisco Unified Communications Manager Express Endpoints 3-137
Rebooting Commands 3_138
Verifying Cisco Unified Communications Manager Express Endpoints 3-139
Verifying Phone VLAN ID 3-140
Verifying Phone IP Parameters 3-141
Verifying Phone TFTP Server 3"142
Verifying Firmware Files 3-143
Verifying TFTP Operation 3"144
Verifying Phone Firmware 3-145
Verifying SCCP Endpoint Registration 3-146
Verifying SIP Endpoint Registration 3"147
Verifying SCCP Registration Process 3-148
Verifying SIP Registration Process 3-149
Verifying Endpoint-Related Dial Peers 3-150
Summary 3-151
Module Self-Check 3-155
Module Self-Check Answer Key 3"160
Dial Plan Implementation 4zl
Overview 4-1
Introducing Call Routing .4-3
Objectives 4"3
Introducing Numbering Plans 4-4
North American Numbering Plan 4-5
NANP Numbering Assignments 4-6
European Telephony Numbering Space 4-9
Fixed and Variable-Length Numbering Plan Comparison 4-11
E.164 Addressing 4-12
Scalable Numbering Plans 4-13
Nonoverlapping Numbering Plan 4-14
Scalable Nonoverlapping Numbering Plan Considerations 4-15
Overlapping Numbering Plans 4-16
Overlapping Numbering PlanExample 4-17
Scalable Overlapping Numbering PlanConsiderations 4-18
Privateand Public Numbering Plan Integration 4-19
Privateand Public Numbering Plan Integration Functions 4-20
Private and PublicNumbering Plan IntegrationConsiderations 4-21
Number Plan Implementation Overview 4-22
Private Number Plan Implementation Example 4-23
PublicNumber Plan Implementation 4-24
Call Routing Overview 4-25
Call Routing Example 4-26
Summary 4-27
Understanding Dial Plans _ 4-29
Objectives 4-29
Defining Dial Plans 4-30
Dial Plan Implementation 4-31
Dial Plan Requirements 4-32
Endpoint Addressing 4-33
Endpoint Addressing Considerations 4-34
Call Routing and PathSelection 4-35
Path Selection Example 4-36
PSTN Dial Plan Requirements 4-37
Inbound PSTN Calls 4-38
2010CiscoSystems, Inc. Implementing CiscoVoice Communications and QoS (CVOICE) v8.0 iii
Numbers in Inbound PSTN Calls 4-39
Outbound PSTN Calls 4-40
Numbers in Outbound PSTN Calls 4-41
PSTN Backup 4-42
ISDN Dial Plan Requirements 4-43
ISDN Dial Plan Requirements 4-44
Inbound ISDN Calls 4-45
Digit Manipulation 4_46
Calling Privileges 4-47
Calling Privileges Example 4-48
Call Coverage 4-49
Call Coverage Features 4-50
Summary 4-51
Describing Digit Manipulation 4.53
Objectives 4.53
Digit Collection and Consumption 4-54
Cisco Unified Communications Manager Express Addressing Method 4-55
User Input on SCCP Phones 4-56
SCCP Digit Collection 4-57
SIP Digit Collection (Simple Phones) 4-58
SIP Digit Collection (Enhanced Phones) 4-59
Dial Peer Management 4-60
Digit Consumption 4-61
Components of Digit Manipulation 4-62
Digit Manipulation Order 4-63
Digit Stripping 4-64
Digit Forwarding 4-65
Digit Prefixing 4-66
Simple Digit Manipulation Comparison 4-67
Number Expansion 4-68
Basic Digit Manipulation Example 4-69
Caller IDDigits Manipulation 4-70
Caller ID Manipulation Example 4-71
Digit Manipulation Using Voice Translation Rules and Profiles 4-72
Voice Translation Rules and Prof les Hierarchy 4-73
Voice Translation Rule Regular Expressions 4-74
Voice Translation Rule Operations 4-75
Prepending Digits 4-76
Voice Translation Rule Search-and-Replace Examples 4-77
Assigning Rules to Prof les 4-79
Translation Profile Processing Order 4-80
Incoming PSTN Call Example 1 4-81
Incoming PSTN Call Example 2 4-82
Incoming PSTN Call Blocking Example 4-83
Digit Manipulation Using dialplan-pattern Command 4-84
Digit Manipulation Using dialplan-pattern Command Example 1 4-85
Verifying Digit Manipulation 4-88
Verifying Dial Plan 4-89
Verifying Translation Rules and Profiles 4-90
Testing Translation Rules 4-91
Summary 4-93
Configuring Path Selection . i^
Objectives 4-95
Call Routing and Path Selection 4-96
Dial-Peer Matching 4-97
Dial Peer Matching Refresher 4-98
Matching Dial-Peer Commands Refresher 4-100
Outbound Dial-Peer Matching Order 4-102
Best Practices 4-13
Path Selection Strategies 4-1 4
Site-Code Dialing andToll Bypass 4-105
Site-Code Dialing and Toll Bypass Example 4-106
Site-Code Dialing and Toll Bypass with Backup Example 4-107
Configuring Site-Code Dialing and Toll Bypass 4-108
Step1:Configure Voice Translation Rules andProfiles for VoIP Intersite Routing 4-109
Step2: Define Dial Peersfor VoIP Intersite Routing 4-110
Step3: Configure Voice Translation Rules and Profiles for PSTN Intersite Routing 4-111
Step4: Define Dial Peersfor PSTN Intersite Routing 4-112
Outbound Site-Code Dialing Example 4-113
Inbound Site-Code Dialing Example 4-114
Tail-End Hop-Off 4-115
TEHOScenario 4-116
Configuring TEHO 4-117
TEHO Configuration Example 4-118
Summary 4-119
Configuring Calling Privileges 4-121
Objectives 4-121
Calling Privileges Characteristics 4-122
Dial-Peer Granularity 4-123
Implementing Calling Privileges onGateways 4-124
COR Elements 4-126
COR Logic 4"127
COR Implementation Example 4-129
Implementing Calling Privileges on SRSTand Cisco Unified Communications
Manager Express 4-130
Calling Privileges Implementation at a Glance 4-131
Configuring COR 4-132
Configuring COR 4-133
Assigning COR 4-134
Configuring COR Labels and Lists (Steps 1 and2) 4-135
Assigning COR Lists to Dial Peers (Step3) 4-136
Assigning CORLists to Endpoints (Step 4) 4-137
Configuring COR for SRST 4-138
Verifying COR 4-139
Verifying CORNames and Lists 4-140
Verifying Dial Peer CORSettings 4-141
Verifying SIPEndpoint COR Settings 4-142
Summary 4-143
ModuleSummary 4-145
ModuleSelf-CheckAnswer Key 4-153
i 2010CiscoSystems. Inc. Implementing CiscoVoice Communications and OoS (CVOICE) v8.0
Module 3
Manager Express Endpoints
Implementation
Overview
This module describes the basic functionality of Cisco Unified Communications Manager
Express This information includes the configuration of specific network components and
services that are necessary for the proper functioning ofCisco Unified Commun.cations
Manager Express.
The module describes features for abasic Cisco Unified Communications Manager Express
system. Manv ofthe features that this module presents arc necessary for asuccesstul
deployment ofCisco Unified Communications Manager Express.
The endpoints that are supported by Cisco Unified Communications Manager Express include
ctco Unified IP phones Sling either Skinny Client Control Protocol (SCCP) or Session
initiation Protocol (SIP). The module describes different types of endpoints, their models, and
capabilities.
Finallv the module explains how to configure the systemwide and endpoint-specific
components of Cisco Unified Communications Manager Express. Special attention is g.ven to
the various types of directory numbers that play akey role in making calls.
Module Objectives
Upon completing this module, you will be able to describe how to implement IP phones using
Cisco Unified Communications Manager Express. This ability includes being able to meet
these objectives:
. Describe the functions and operation ofthe Cisco Unified Communications Manager
Express
Describe all components that are required to support endpoints by Cisco Unified
Communications Manager Express and explain how to configure them
Describe Cisco Unified Communications Manager Express endpoint configuration
elements such asphones and directory numbers
3-2 Implemenling Cisco Voice Communications and QoS (CVOICE) v8~ ' 2010 Cisco Systems, Inc
Lesson 1
Introducing Cisco Unified
Express
Overview
Cisco Unified Communications Manager Express provides call processing for Cisco Unified IP
phones for small-office orbranch-office environments. It enables the large portfolio ofCisco
integrated services routers todeliver unified communications features that are commonly used
by business users to meet the voice and video communications requirements ofthe small or
medium-sized office. CiscoUnifiedCommunications ManagerExpress allows the deployment
ofacost-effective, highly reliable communications system using asingle device with Cisco
IOS Software.
This lesson introduces die key features and functionality of Cisco Unified Communications
Manager Express and explains what is required todeploy it onCisco IOS routers.
Objectives
Upon completing this lesson, you will be able to describe the functions and operation ofthe
Cisco Unified Communications ManagerExpress, 'fhis abilityincludesbeingable to meet
these objectives:
Describe the functionsofthe Cisco Unified Communications Manager Express in a voice
network
Identify the key features and benefits ofCisco Unified Communications Manager Express
Describe the supported platforms and the required memory, licensing, and software that is
needed to deployCiscoUnifiedCommunications ManagerExpress
Explain the operation oftheCisco Unified Communications Manager Express when calls
are made between the Cisco Unified Communications Manager Express and PSTNand
between two Cisco Unified Communications Manager Express servers
Introducing Cisco Unified Communications
Manager Express
This lopie provide;, an overview of Cisco Unified Communications Manager Express and its
deployment.
3-4
Cisco Un
ified Communications
Manager Express Overview
mmmmm Function H'tesrnwirai m ^v.^r^.mm^mm
Call processing Routing, originating, aid tenninating calls, including
statistical collection.
Signaling and
device control
Signals connections between call endpoints. Makes
devices such as phones and gateways (o establish and
tear down streaming sessions.
Dial plan
administration
Dial plan consists of dial peers an<f registered phone
numbers. It determines call routing.
Phone feature
administration
Provides services such as hold, transfer, forward,
conference, speed dial, last-number redial, and call park to
Cisco Unified IP phones and gateways.
Directoryservices Maintains its own database lo store user- and phone-
related informati on.
Direct access to
galeway features
and modules
1n-the-box access to DSP resources and network module
functions (such as voice mail).
CiscoUnified Communications Manager Express extendsenterprise telephony features and
functions to packet telephony nctvvork devices. These packet telephony networkdevices
include Cisco IPphones, media-processingdevices. VoIP gateways, and multimedia
applications.
Cisco UnifiedCommunications Manager Express provides these functions:
Call processing: Call processingrefers lo the complete process of routing, originating, and
terminating calls, including any billing and statistical collection processes,
Signaling and device control: Cisco UnifiedCommunications Manager Expresssignals
calls between endpoints and directs devices such as phones, gateways, and conference
bridges to establish and tear down streaming connections.
Dial plan administration: The dial plan is a set of dial peers that Cisco Unified
Communications Manager Express uses to dctennine call routing. Cisco Unified
Communications Manager Express provides the ability to create scalable dial plans.
Phone feature administration: Cisco Unified Communications Manager Express offers
services such as hold, transfer, forward, conference, speed dial, last-number redial. Call
Park, and other features to Cisco Unified IP phones and gateways.
Director} services: Cisco Unified Communicalions Manager Express stores user- and
phone-related data in the NVRAM ofthe Cisco IOS router.
Direct access to gateway features and modules: Cisco Unified Communications Manager
Express runs on the Cisco IOS router and lias direct access to the digital signal processor
(DSP) resources and modules that are installed in it.
Implementing Cisco Voice Communications and QoS (CVOICE) vfi 12010 Cisco Systems, Inc.
Cisco Unified Communications Manager Express Positioning
The figure illustrates the positioning ofCisco Unified Communications Manager Express
within the Cisco Unified Communications portfolio.
Manager Express Positioning
Cisco Unified
Communications
Applications
Call Processing
Add network infrastructure as required
Ability to add services, collaboration,
messaging, customer contact, etc.
Cisco offers four different product options for call processing as follows:
CiscoSmart Business CommunicationsSystem: Thisproduct runson theCiscoUnified
Communications 500 Series for Small Business platformand supports up to 104 users.
CiscoUnified Communications Manager Express: This platform runs on the Cisco
Integrated Services Routers andoffers support for up to365users.
Cisco I'nified Communications Manager Business Edition: This softwareproductruns
on the Cisco 7800Series MediaConvergence Serversand supportsup to 500 users.
Cisco Unifled Communications Manager: This softwareproduct runs on the Cisco7800
Series Media Convergence Servers or theCiscoUnified Computing System. The Cisco
Unified Computing System provides thehighest scalable platform thatradically simplifies
thetraditional architectures, reducing thenumber of devices that mustbe purchased,
cabled, configured, powered, cooled, andsecured. Thesolution delivers end-to-end
optimization for virtualized environments while retaining theability tosupport traditional
operating system and application stacks inphysical environments. It is well-suited for the
largest Cisco Unified Communications Manager deployments for upto30,000 and more
users.
Cisco Unified Communications Manager Express Endpoints Implementation
Cisco Unified Communications Manager Express Deployment
Models
Ihis subtopic describes the twodeployment models of Cisco Unified Communications
Manager Express.
Manager Express Deployment Models
router with optionally installed Cisco Unity Express
network module
Internet
PSTN
Cisco Unified
Communications
Manager Express/
^Cisco Unity Express
Single-site deployments use the public switched telephone network (PSTN) eonimunications
for all off-site voice traffic. One Cisco Unified Communications Manager Express site supports
up to 365Cisco Unified IPphones. If the Cisco Unity Express module is installed inthe router,
voice-mail sen ice is available.
Cisco Unified Communications ......
Express Deployment Models (Cont/
Multiple Site Deployment
Cisco unified messaging gateway acts as a central hubforrouting
messagesandexchanging subscriber anddirectory information within .
unified messaging network.
Cisco Unified Communications |
Manager Express / Cisco Unity^j
Express
Cisco Unified
Communications
Manager Express/
Cisco Unity Express
Unified Messaging Gateway
Multiple Site Deployment
Multiple site deployments place VoIP calls between sites. When the 11.323 protocol isused for
communications between clusters, anH.323 gatekeeper can beused for call routing and Call
Admission Control (CAC). Remote sitescanbe Cisco Unified Communications Manger
clusters or Cisco UnifiedCommunications Manager Express sites.
When voice mail networking is required, a Cisco Unified Messaging Gateway provides a
centralized Voice Profile for Internet Mail (VPIM) routing andresolution service. Thisservice
routes callsbetween voice-mail systems by using Simple Mail Transfer Protocol (SMTP) to
deliver voice mail that was recorded at the source, adding the message as an attachment to an
email message that issent tothedestination. The Cisco unified messaging gateway
synchronizes its local database with all the registered voice-mail systems tocreate a global
voice-mail directory, such as a Domain Name System (DNS). Any user wishing tosend the
same voice mail to people located inmultiple sites looks uptherecipients inthe global
director)' and assigns them asneeded toa single voice-mail message. The message isthen
relayed through the Cisco unified messaging gateway toall therecipients, without placing a
single external phone call.
Cisco UnifiedCommunications Manager Express Endpoints Implemenlation 3-7
Key Features and Benefits
This topic describes the key features and benefits ofCisco Unified Communications Manager
Express.
3-8
Key Benefits
Supports deployments of up to 365 phones on a single router
* Extends a rich set of enterprise capabilities to the small office
Delivers voice, video, and data over a consolidated
infrastructure
Is based on Cisco IOS Software
* Supports converged applications
* Can be administered by GUI or CLI
Cisco Unified Communications Manager Express is a feature-rich, entry-level IPtelephony
solution that is integrated directly into Cisco IOS Software andcanscale upto 365 phones.
Cisco UnifiedCommunications Manager Express allows small- to medium-si/cd businesses
andautonomous small enterprise branch offices todeplov voice, data, andIPtelephony on a
singleplatform, therefore streamlining operations and lowering network costs.
Cisco UnifiedCommunications Manager Express is ideal for customers who hovedata
connectivity requirements and have a need for a telephony solution in the same office. Whether
offered through themanaged service offerings of a service provider or purchased directly by a
corporation. Cisco UnifiedCommunications Manager Express offers most ofthe core
telephony features required in the small office, and many advanced features not available with
traditional telephonv solutions. Theability lodeliver IPtelephony anddatarouting byusing a
singleconverged solution allowscustomers to optimizetheir operations and maintenance costs,
resulting in a vcry cost-effective solution that meets office needs.
Because the solution is basedon Cisco IOS Software, it buildson convergent networks that
include content networking, video, quality of service (QoS). firewall, and XMEservices.
Administrationand management are accomplished through either the familiar Cisco IOS
Software command-line interface (CLI) or a web-based GUI.
Implementing Cisco Voice Communicalions and QoS (CVOICE] vS.O
2010 Cisco Systems. !nc
Phone Features
This subtopic provides an overview of key phone features.
Phone Features
Endpoints
Analog phones and
fax machines
SRTP
Extension Mobility
XML Services
Directory Lookup
Dial Enhancements
Call handling
Single Number
Reach (SNR)
AI Cisco Unified IP phone models. Do* single-line and mulHIine.
FXS ports and the CiscoAnalog Telephone Adaptor (ATA) 166.
End-tc-end vblP security between Cisco Unified !Pphones.
User extension applied to Unified IP phone when the user logs in.
Cisco Unified IP phones allow access tothlr*party applications,
using phone display.
Browsing through local user database.
Speed dial, last-number dial, on-hoc* dial.
Call transfer(consultative, blind), caH held and retrieve, pickup of
held calls, call waiting.
Calls tothe enterprise number simultaneously ring adesk set and i
cell phone and can be answered ateither. Calls can be switched
from cell phone to IP phonewith onefcuttofl press. Desk phone
number can besent ascaller ID instead oforiginal calling number.
The following are high-level phone features of Cisco Unified Communications Manager
Express:
> Support for the complete line of Cisco single-line and multiline IP phones
. Support for analog phones and fax machines on the Cisco Unified Communications
Manager Express router analog voice ports and on the Cisco Analog Ielephone Adaptor
186 (ATA 186)
Media encryption using Secure Real-Time Transport Protocol (SRTP)
Cisco Extension Mobility
XML services on Cisco IP phonesXML-based directory services
Call handling:
On-hook dialing
Speed dial and last-number redial
Call Transferconsultative andblind
Call hold and call retrieve
Call Pickup of on-boldcalls
Call waiting
Tone onhold and tone ontransfer forinternal calls
Local directory lookup
Configurable ringtypes
Do Not Disturb (DND) feature to divert calls directly to voice mail
3-9
Single Number Reach (SNR,: Calls to the enterprise number simultaneously ring adesk se
and acell phone and can be answered at either. Calls can be switched from cell phone
phone with one-button press. Desk phone number can be sent as caller ID instead of
original calling number.
t
to IP
3-10 Implementing Cisco Voice CommunicationE and QoS (CVOICE] vB 0 20,0 Cisc0 Syslems |nc
KjMt
System Features
This subtopicprovidesan overview of key systemfeatures.
System Features
WM^mm fertnr-* Description ^H
CU. GUI. Cisco
Configuration
Professional
Administration
CUor GUIadministration(GUI has less granularity).Cisco
Configuration ProfessionalIs a free downloadable tool to reduce
the configuration time.
Survivable
Remote Site
Telephony
(SRST)
Telephony backup services to help ensure that the branch office
has continuous telephony service. Cisco UnifiedCommunications
Manager Express takes over the role of the Cisco Unified
Communications Manager during IP connectivity loss.
Music on HoW
(MOH)
Multiple local files on router or external livesource,
Class of
Restriction
Control of call permissions.
Conferencing Multiple participants in one calf.
Auto Attendant Interactive voice response (IVR)AutoAttendant.
CTI Interface Industry-standard computer telephony integration(CTI)interface
allows interoperability with third-party products.
Reporting Call Detail Record (CDR) generalion via RADIUS.
The following are high-level systemfeatures of CiscoUnifiedCommunications Manager
Express:
Multiple administration methods:
CLI
Web-based embedded GUI for moves, adds, and changes
CiscoConfiguration Professionalan administrator tool that helps reducethe
configuration time
Cisco Unified Survivable Remote Site Telephony (SRST): Telephony backup services to
ensure that the branch office has continuous telephony service. Cisco Unified
Communications Manager Express takes over the role ofthe Cisco Unified
Communications Manager during IP connectivity loss.
Signaling encryption
Hardware and software conferencing capabilities
Music on hold (MOH): Played on transfer of external calls. Multiple local tiles on the
router or external feed source.
Paging. Intercom
Distinctive ringinginternal versus external
International language support
Cisco Unified IP Interactive Voice Response (IVR) Auto Attendant
Class of restriction to restrict calling capabilities
i 2010 Cisco Systems, Inc. Cisco Unified Communications Manager Express Endpoints Implementation 3-11
Computer telephony integration (CTI) support with Cisco Telephony Application
Programming Interface (TAPI) Lite
Call Detail Record (CDR) generation via RADIUS
3-12 Implementing Cisco Voice Communications and QoS (CVOICE] v8.0 2010 Cisco Systems. Inc
JM>
Trunk Features
This subtopic provides an overview ofkey Cisco Unified Communications Manager Express
trunk features.
Trunk Features
^^^1 PAAttlfP Dfiscriotion ^^H
Trunk Support Digftsd T1/E1, ISDN PRI/BR1, analog FXS,
FXO, FXS-DID, E&M
DIDandDOD Direct inward dialing (DID)and direct outward
dialing (DOD)
CLID
Caller identificationdisplay and blocking,
callingname display, and automaticnumber
identification support
MappingTrunk-to-Phone Dedicated trunk mapping to phone button
Transcoding Use of DSP nasourcesfortranscoding
Conferencing Multiple participants in one call
H.323 and SIP Support Interoperability withH.323and SIP signaling
gateways
The following are high-level trunk features of Cisco Unified Communications Manager
Express:
Direct inwarddialing(DID) and direct outwarddialing(DOD)
BRI andPRI supportall switch typesthat Cisco IOSSoftware supports
Caller identification display and blocking, calling name display, and Automatic Number
Identification (ANI) support
AnalogForeignExchange Office (FXO). DID
Digital trunk supportTl and El
WAN linksupportFrame Relay.AIM, Multilink PPP(MLP), and DSL
Network calls using H.323
Dedicated trunk mapping to phone button
H.323 to Session Initiation Protocol (SIP) call routing to Cisco Unity Express
RFC 2833 support over SIP trunks
Transcoding
) 2010 Cisco Systems, Inc. Cisco Unified Communications Manager Express Endpoints Implementation 3-13
Voice-Mail Features
This subtopic prov ides anoverview of key Cisco Unified Communications Manager Express
voice-mail features.
Voice-Mail Features
Integration with Cisco Standalone voice-mail server platform used in
Unity large enterprises
Integration with Cisco Module-based integrated voice-mail solution
Unity Express for small enterprises or branch offices
Integration with third- H.323 or DTMF-based integration with third-
party voice mail party voice-mail systems
Integration with Cisco Routing of voice-mail messages and
unified messaging exchanging subscriber and directory
gateway information within a unified messaging
network
Voice-mail enhancements Fast voice-mail access, Message Waiting
for Cisco Unified IP Indicator (MWI)
phones
The following are high-level voice-mail features for Cisco Unified Communications Manager
Express:
Integration with Cisco Uniiv voice mail
Integration with Cisco Unit)' Express voice mail
Third-party voice-mail integration11.323. analog dual tone multifrequency (DTMF)
Integration v\ itli Cisco unified messaging gateway -routing of voice-mail messages and
exchanging subscriber and director} infonnation within a unified messaging network
Voice mail enhancements for Cisco Unified IP phonesfast voice mail access. Message
Waiting Indicator (MWI)
3-14 Implementing Cisco Voice Communicalions and OoS (CVOICE| v8 0 ) 2010 Cisco Systems. Inc
#*
"mm
^Mtf.
Supported Platforms
This topic describes the platforms that support the Cisco Unified Communications Manager
Express.
Supported Platforms
Cisco Unified Communications Manager Express is
supportedon these Ciscoplatforms:
Cisco 1861 Integrated Services Router
Cisco 2800 Series routers
Cisco 2900 Series routers (G2)
Cisco 3800 Series routers
Cisco 3900 Series routers (G2)
Cisco Unified Communications Manager Express supports these Cisco platforms:
Cisco 1861 Integrated Services Router
Cisco2800. 2900. 3800. and 3900 Series Integrated Services Routers
Note Cisco 2900 and 3900 Series Integrated Services Routers arereferred toas Generation 2
(G2) router platforms.
Cisco Unified Communications Manager Express Endpoinls Implementation
3-15
Cisco Integrated Services Routers Scalability
This subtopic describes the capabilities that are offered by the Cisco 1861, 2800 Series, and
3800 Series Integrated Services Routers.
Cisco 3825 and 3845 platforms allow ahigher number ofphones that are supported in SRST
mode than in regular Cisco Unified Communications Manager Express mode. The SRST mode
is enabled onlv during WAN failures, when branch phones lose IP connectivity to the Cisco
Unified Communications Manager cluster and fall back to the local SRST gateway.
The modular chassis ofthe Cisco 2800 and 3800 Series Integrated Services Routers allow the
in-the-box installation ofadditional network modules, most notably the Cisco Unity Express
module that provides voice mail features.
3-16 Implementing Cisco Voice Communications and QoS(CVOICE) vS.O
Cisco Integrated Services Routers Generation 2Scalability
This subtopic describes the capabilities that are offered by the Cisco 2900 and 3900 Series
Integrated Services Routers.
3i
Cisco Integrated Services Router;
Generation 2 Scalability
Cisco Unified Communicalions Manager Express
Cisco Umf-edSurvivable Remote Site To,or,hn'11'
100/(00
Phones
iltiple Services
Ertended Modular
Connectpvity (EVM
ISM SM WICA/lc)
400-1100
myn-usually ucin-vS
Modularity witti Performance
Optimizedtor "All-in-Ors"
Solution (NM-SM, NME,
EVM ISM, WICA/IC)
Cisco Unity Express
Local Auto MTerxtant and Voice
MailSystem witt 12-250
Mailtxwes. 4-24 Sessions,
300 Hours ot Storage
Cisco 3925 and 3945 platforms support ahigher number ofphones in SRST mode than in
regular Cisco Unified Communications Manager Express mode.
The modularity ofboth series enables an easy integration with additional gateway features.
Cisco UnifiedCommunications Manager Express Endpoirtis Implementation
3-17
Memory Requirements
This subtopic lists the memory requirements for Ihe Cisco Unified Communications Manaacr
Express platforms.
Memory Requirement
Platform
!Number of Phones 'TRAM (Mitt
rmm
Cisco 1861 25
Cisco 2801 25
Cisco 2811 35
Cisco 2621 50
Cisco 2901 42
Cisco 2911 58
Cisco 2921 110
Cisco 2951 165
Cisco 3825 175
Cisco 3845 150
Cisco 3925 250
Cisco 3925E 400
Cisco 3945 350
Cisco3945E 450
256 128
256 128
256 128
256 128
512 256
512 256
512 256
512 256
364 128
384 128
1024 512
1024 512
1024 512
1024 512
fhis table lists die memory that isrequired by each Cisco Unified Communications Manager
Express platform. The number of supported phones represents the highest number onthegi\en
platform. Ifa lower number ofphones is needed in an enterprise environment, the router may
perform well with less RAM. but the provided memory figures arehighly recommended.
3-18 Implementing Cisco VoiceCommunications and QoS (CVOICE) v8.0
Cisco Integrated Services Routers Licensing and Software
This subtopic describes the license and software requirements ofthe Cisco Unified
Communications Manager Express.
Cisco Integrated Services Routers
Licensing and Software
* Right-to-use licensing:
- Licensed Cisco IOS Software
* Cisco IOS Release 15.0.1MforCisco Unified
Communications Manager Express 8.0
IP voice feature set
- Feature license based on maximum number of phones
- Phone license per phone
Cisco Unified Communications Manager Express additional
software and files
- GUI files (optional)
- Firmware for chosen endpoint models
Cisco Unified Communications Manager Express uses therighl-to-usc licensing approach, in
which theCisco Unified Communications Manager Express feature license entitles an
enterprise to use the feature. This license is based on the number ofendpoints to be deployed.
Each Cisco Unified IP phone orCisco ATA port requires aCisco Unified Communications
Manager Express seat license.
The following are the requirements for a Cisco Unified Communications Manager Express
release on a supported router:
Cisco IOS Software Release 15.0.1M or greater
I!1 voice feature set for Cisco IOS Software
The appropriate amount offlash memory and RAM in the router
You must download and configure additional files ifyou want to use the optional GUI interface
or the Cisco Configuration Professional. These tools are not covered inthis course.
Youmustdownload andinstall thefirmware files for themodels of phones that youchoose to
deploy with Cisco Unified Communications Manager Express. You can retrieve these files
from http://ww-w.cisco.com. The download for these files and the GUI files requires aCisco
login.
Cisco UnifiedCommunications Manager Express Endpoints Implementslion
Cisco Integrated Services Routers Generation 2Licensing
Model
This subtopic describes the license and software requirements of the Cisco Unified
Cisco integrated Services Router;
~ 'ion 2 Licensinq Model
Software activation licensing(planned)
Right-lo-jse licensing (current)
Entrylevel25user seats across all platfomis
Integrated ServicesRouters-based licenses transferable to Integrated
Services Routers Generation 2
CiscoUnrfied Communications ManagerExpress and SRSTcount licenses
interchangeable within the same number of user counts
âi-nrimniii
Platform
Platform
Counted for x phones
FL-CME
FL-SRST
FL-CME-SRST-x
Unified CME-SRST
C29xx-CME-SRST/K9
C39xx-CME-SRST/K9
PVDM3
SL-29-UC-K9 or SL-39-UC-K9
FL-CME or FL-SRST and
FL-CME-SRST-25
Cisco Generation 2 platforms (Cisco 2900and3900 Series Integrated Services Routers)
introducea new licensing approach that uses license-basedsoftware activation. A universal
Cisco IOS image is combined with multiple package options. ThenewCisco Unilied
Communications Manager Express and SRST bundles for the 02 routers provide the entrj level
for 25 user seats across all platfomis. The bundles include unified communications technology
packets. Hash, and DRAM. Currently, the software activation license approach isnot yet
implemented for Cisco Unified Communications Manager Express or SRST. Forthese features,
the old-style, owner-based, right-to-use licensing approach isstill inplace, fhe licensing for
Cisco Unified Communications Manager Express andSRST is interchangeable within thesame
number of user counts, for investment protection purposes. The licenses can be transformed
from a Cisco Integrated Services Router-based platform intoa Cisco Integrated Services Router
G2-based plat fonn.
The figure illustrates the difference between Cisco Integrated Services Router-based and Cisco
Integrated Sen ices Router G2-bascd licensing. TheCisco Integrated Services Router license
includes two components: the piatform-related Cisco Unified Communications Manager
Express or SRST feature license, and the per-seat feature license (either Cisco Inified
Communications Manager Express or SRST). The CiscoIntegrated Serv ices Router Ci2
package is related to theseries (Cisco 2900 or 3900 Series) andincludes three components:
Packet \oiec DSP module (PVDM) license.
Cisco Unified Communications license.
License touse either Cisco Unified Communications Manager Express or SRS'l. including
25 user seats. Additional per-seat licenses must be purchased separately.
Implementing Cisco Voice Communications and QoS (CVOICE) vS.O
A^^prt
Pm
Operation
This topic describes the operation of Cisco Unified Communications Manager Express.
Calls to and from PSTN
1001
P phone side:
Virtual dial peers
created automatically
for phone extensions
DNIS 555-1001
555-2001
dirct-inward-dial
I
port 1/0/0:23
The figure illustrates the operation of Cisco Unified Communications Manager Express running
on avoice gateway that is connected to the PSTN over adigital trunk. It has multiple Cisco
Unified IP phones registered to it. The registered phones can make calls to each other. The calls
are signaled by exchanging messages between the phones and the Cisco Unified
Communications Manager Express, but the media flows directly between the phones. The
gatewav routes calls to external destinations over the dial peer that uses the T1 channelized
controller. When making and receiving calls to and from the PSTN, the gateway typically
performs digit manipulation in the calling (ANI) and called numbers (Dialed Number
Identification Service [DNIS]). With this approach, numbers are made routable in the PSTN
andareshortened to internal numbers within theenterprise network.
Cisco Unified Communications Manager ExpressEndpoinls Implementation 3-21
Calls Through IP Network
This topic describes the operation of two Cisco Unified Communications Manager Express
platforms that are located in the branches ofan enterprise.
Calls Through IP Networf
1001
Cisco Unified
Communications
Manager Express
10 111
Cisco Unified
Communications
Manager Express
10222
2001
Virlua dial peers for pnones
1001 1002 1003
dial peer voice 1 v oip
deac nati on-pattern 2
seas on-t arget 10.2 2.2
Internet
Virtua dial peers Forph mes.
2001 2002, 2003
dial peer voice 1
v
5ip
dest nation-patte 1...
seas on-target 10 1 1.1
Call routing isbased onthe dial peer matching. Two hpes ofdial peers are involved in this
scenario, as follows:
Plain oldtelephone service (POTS) dial peers: POTS dial peers, also referred toas
virtual dial peers, are automatical!) created on the Cisco Unified Communications Manager
Express for directory numbers onregistered phones. In this way. the Cisco Unified
Communications Manager Express with address 10.1.1.1 creates three POTS dial peers for
numbers 1001. 1002. 1003. The Cisco Unified Communications Manager Express 10.2.2,2
has dial peers for numbers 2001. 2002. and 2003.
VoIP dial peers: The VoIP must be configured by the administrator toallow calling
between both sites.
3-22 Implementing CiscoVoice Communications and QoS (CVOICE] v80
Summary
Summary
Cisco Unified Communications Manager Express is a Cisco
IOS Software-based call processing platformtargeted at
small-to-medium enterprises and branch offices.
It offers an extensive set of phone, system, trunk, and voice-
mail features.
Cisco Unified Communications Manager Express is
supported on Cisco Integrated Services Routers (1861, 2800,
3800)and CiscoIntegrated Services Routers Generation 2
(2900, 3900) and is currently licensedon the right-to-use
basis.
' Forconnectivity over VoIP and POTSnetworks, the Cisco
Unified Communications Manager Express uses virtual dial
peersfor locally registered Cisco Unified IPphones, dial
peers for external destinations, and digit manipulation.
Cisco UnifiedCommunications Manager Express Endpoints Implementation
3-24 Implementing Cisco VoiceCommunications and QoS (CVOICE) u8.0 201C Cisco Systems, Inc
Lesson 21
Examining Cisco Unified
Express Endpoint
Requirements
Overview
It is important to be able to distinguish between various Cisco Unified Communications end-
user devices that you may encounter during the course ofdeploying and administering aCisco
Unified Communications network. In addition, understanding the boot and registration
communication between a Cisco Unified IPphone and theCisco Unified Communications
Manager Express is critical for understanding normal voice network operations and for
troubleshooting. This lesson introduces the endpoints that are supported by Cisco Unified
Communications Manager Express anddescribes their features.
Objectives
Upon completing this lesson, you will be able todescribe the endpoints that can interoperate
with Cisco Unified Communications Manager Express and thecomponents thatarerequired to
support the endpoints, such as Cisco Discovery Protocol, TFTP, DHCP, Network Time
Protocol (NTP), and their configuration. This ability includes being able tomeet these
objectives:
Describe the Cisco Unified Communications Manager Express 8.0SCCP andSIP
endpoints andexplain theircapabilities
Explain the Cisco Unified IP phone endpoint boot process and identify its major
requirements (PoE, VLAN, DHCP, TFTP)
Describe options topower endpoints and describe their characteristics
Describe endpoint VLAN requirements; explain voice and data VLANs and how to
configure VLANs to enable endpoint registration
Identitj DHCP service options and DHCP relay, and describe how to configure them to
support Cisco Unified Communications Manager Express endpoints
Describe VI Pand howto configure it
Describe Cisco Unified IP phone firmware files and XML configuration files, and identify
howCisco Unified IPphones obtainthe files via TFTP
Describe how toconfigure s\ stem-level parameters inan SCCP environment
Describe how to sel up system-le\el parameters ina SIPenvironment
Overview of Cisco Unified Communications
This topic introduces Cisco Unified Communications Manager Express endpoints, describes
their capabilities, and explains the signaling protocols that they can operate with.
Overview of Endpoints
Cisco SCCP-Only
Phones
Analog Station Gateways
SCCP Video
Phones
Third-Party
SIP Endpoints
Avariety ofendpoints, Cisco products as well as third-party products, can be used with Cisco
Unified Communications Manager Express. The endpoints include Cisco Unified IPphones,
analog station gateways (which allow analog phones tointeract with Cisco Unified
Communications ManagerExpress), and videoendpoints.
Cisco Unified Communications Manager Express supports two protocols to beused for
endpointsSkinny Client Control Protocol (SCCP) and Session Initiation Protocol (SIP).
3-27
Endpoint Signaling Protocols
This subtopic describes the signaling protocols that are used by the Cisco Unified
Communications Manager Express endpoints.
Endpoint Signaling Protocols
supports endpoints using SCCP and SIP:
Cisco proprietary SCCP:
OnlyCisco Unified IPphones
Largest set oftelephony features
- Standard SIP:
Standard compliant third-party phones
Basic telephony features
StandardSIP with CiscoUnified Communications Manager
Express extensions:
Only Cisco Unified IP phones
Rich set oftelephony features, depending on phone
model
From a feature support perspcclhe. the protocols can hecategorized into three groups as
follows:
SCCP: SCCP isa Cisco proprietary' protocol and typically only used by Cisco Unified IP
endpoints. SCCP offers a rich set oftelephony features, most of which aresupported onall
Cisco Unified IP phone models.
Standard SIP: Cisco Unified Communications Manager Express supports standards-based
SIP endpoints. Thenumber of standardized telephony features, however, is limited
compared to feature-rich SCCP.
Cisco Unified Communications Manager Kxpress SIP support for Cisco Unified IP
phones: \\ henCisco Unified Communications Manager Express interacts withCisco
Unified IPphones using the SIP protocol, many features aresupported inaddition tothe
standard feature set of SIP. Cisco Unified Communications Manager Express supports
similarfeatures for Cisco Unified IP phones that are supported withSCCP, but the number
of features that aresupported depends ontheparticular model of Cisco Unified IPphone.
Endpoint Capabilities
This subtopic explains the various capabilities that are provided by specific endpoints.
Endpoint Capabilities
CiscoUnified IPphones differ in theircapabilities:
Screen: resolution, size, and color; touchscreen
Codecsupport: G.729, G.711, iLBC, wideband
LAN: speed, PC port
Buttons, navigation dusters
Speakerphone and headset support
Number of lines
Special features: video, conference station, Wi-Fi
Cisco Unified IP phones cover awide range oftypes, from simple, display-less, entry-level
phones to upper-level phones with high-resolution, color, touchscreen displays. DilTerences in
hardware-related capabilities include the following:
Screen: Different modelshave screenswithdifferentresolution, size, color, and
touchscreen capabilities.
Codec support: All Cisco Unified IP phones support G.711 and G.729 codecs. High-end
models also support Internet Low Bitrate Codec (iLBC) and wideband codecs for superior
voice quality.
LAN: Most IPphones have a PC port sothat a PC can beconnected tothe network without
requiring its own wall socket, in-house cabling, and physical switch port. Different phone
models support different speeds on the PC port and on the IP phone switch port (the port
that is connected to a LAN switch).
Buttons, navigation clusters, andsoon: The number of IPphone buttons, softkeys, and
other buttons also differs perphone model. There arealso differences inthetype of
navigation clusters(two-way or four-way).
Speakerphone and headset support: Some IP phones offer speakerphone and headset
support.
Number of lines: The number of lines also differs per phone model.
Other features: Some IPphones provide other special features such as video, Wi-Fi
support, ordedicated support for use inconference rooms (enhanced speakerphone
capabilities, including the option toconnect multiple microphones).
3-29
Basic Cisco Unified IP Phone Models
fhis subtopic describes the basic Cisco Unified IP phone models.
iasic Cisco Unified IP Phone Mot
Basic-featured Cisco IP phones for low-to-medium
telephone use
Single line or directorynumber
Message Waiting Indicator
Cisco Unified IP Phone
7906G.7911G
Cisco Unified IP Ptione
7921G
7931G
fhe basicCiscoUnified IPphones include these models:
CiscoI'nified IP Phones7906 and 7911: These phones fill thecommunication needs of
cubicle, retail, classroom, or manufacturing workers, oranyone who conducts low-to-
moderate telephone traffic, hour dynamic softkeys guide users through core business
features and functions, while a pixel-based display combines intuitive features, calling
information, and XML services into a rich user experience. Both phones offer numerous
important security features, plus the choice of IEEE 802.3af Power over Ethernet (Pol7).
Cisco inline power, or local power through an optional power adapter.
i Cisco Unified Wireless IPPhone 7921: This phone provides a powerful, converged
solution with an intelligent wireless infrastructure. This wireless phone supports a host of
callingfeatures and voice-quality enhancements. Because theCisco Unified Wireless IP
Phone 7921 is designed togrow with system capabilities, features will keep pace with new
system enhancements.
i Cisco I'nified IP Phone 7931: This phonemeetsthe communication needsof retail,
commercial, and manufacturing workers, plus anyone with moderate telephone traffic and
also specific call requirements. Dedicated hold, redial. and transfer keys facilitate call
handling in a retail environment. Illuminated mule and speakerphone keys give a clear
indication ofspeaker status. Apixel-based display with awhite backlight makes calling
infonnation easv to seeanddelivers a rich userexperience.
3-30 Implementing CiscoVoice Communications and QoS (CVOICE) v8
Midrange Cisco Unified IP Phones
This subtopic describes the midrange Cisco Unified IP phone models.
Midrange Cisco Unified IP Phones
Full-featured Cisco Unified IP phones
Multiline
Message Waiting Indicator
. Large pixel-based displays (Cisco Unified IP Phone 7940G and
796 OG)
High-resolution displays for advanced applications
- Cisco Unified IP Phone 7941G. 7942G. 7961G, and7962G
Integrated switches
Built-in headsets and high-qualityspeakerphones
Cisco Unified P Phone
7940G.7941G, 7942G
7960G, 7961G.7962G
Midrange Cisco Unified IP Phones 7940, 7941, 7942, 7960, 7961, and 7962 address the
communications needs ofatransaction-type worker. They provide two or four programmable
line and feature keys, plus ahigh-quality speakerphone. These phone models have four
dynamic softkeys that guide users through call features and functions. Abuilt-in headset port
and an integrated Ethernet switch are standard with these phones. The phones also include
audio controls for the full-duplex, high-quality, hands-free speakerphone, handset, and headset.
The Cisco Unified IP Phones 7941 and 7961 have lighted line keys, and the Cisco Unified IP
Phones 7942 and 7962 add support for the high-fidelity wideband codec.
Note For adetailed list of features per phone model, refer to the data sheets of the Cisco Unified
IPPhone7900Series products.
CiscoUnified Communications ManagerExpress Endpoints Implementation
Upper-End Cisco Unified IP Phones
This subtopic describes the upper-range Cisco Unified IP phone models.
Upper-End Cisco Unified IP Phones
Addresses needs of executives
Large, color, pixel-based displays
*Touch-sensitive display (Cisco Unified IP Phone 7975G)
*Two to eight telephone lines, orcombinations oflines, and
direct access to telephony features
Four orfive interactive softkeys
*Built-in headsets and high-quality, hands-free speakerphones
Cisco Uufled
IP Phone 794SG
CiscoUniHed
P Phone 7965G
CiscoUnified
iPPhnnp7q7sr,
Cisco Unified
P Phone 896t
Upper-end Cisco Iinified IP Phones 7945. 7965. 7970. 7971. and 7975 demonstrate the latest
advances in VoIP telephonv. including wideband audio support, backlit color displays, and an
integrated Gigabit Ethernet port. They address the needs ofexecutives and transaction-type
workers with significant phone traffic, and the needs ofthose working with bandwidth-
intensive applications on collocated PCs.
These IP phones include alarge, baeklit. casj -to-read color display for easy access to
communication infonnation. timesaving applications, and features such as date and time,
calling partv- name, calling partv number, digits dialed, and presence infonnation. They also
accommodate XML applications that take advantage ofthe display. The phones provide direct
access totwo to eight telephone lines (orcombination of lines, speed dials, and direct access to
telephonv features), four or t'lsc interactive softkeys that guide you through call features and
functions, and an intuitive four-vvav (plus Select key) navigation cluster. Ahands-free
speakerphone and handset that is designed for high-fidelity wideband audio are standard, as is a
built-in headset connection.
Note
For a detailed list offeatures per phone model, refer tothedata sheetsoftie Cisco Unified
IP Phone 7900 Series products.
3-32 Implementing Cisco Voice Communications andQoS(CVOICE) v80
Video-Enabled Cisco Unified IP Phones
This subtopic explains the video-enabled Cisco Unified IP phones.
Video-Enabled Cisco Unified IP Phones
Video capabilities (camera, LCD screen, speaker)
- Cisco Unified IP Phone 9900 Series:
- Headset convenience: Bluetooth 2.0or USB wired
headsets
- Tricolor illuminated LED line and feature keys
- XMLandMIDletapplications
Cisco Unified
IP Phone7985G
2
Cisco Umfied
IP Phone 9951
Cisco Unified
IP Phone 9971
Cisco ofTers arange of video-enabled Cisco Unified IP phones that include the following
models:
Cisco Unified IP Phone 7985: This is apersonal desktop videophone for the Cisco Unified
Communications solution. Offering aproductivity-enhancing tool that makes instant, face-
to-face communication possible, the Cisco Unified IP Phone 7985 has avideo call-camera.
LCD screen, speaker, keypad, and ahandset incorporated into one unit.
Cisco Unified IPPhone 9951: This is an advanced collaborative media endpoint that
provides voice, video, applications, and accessories. Highlights include interactive video
with support from the Cisco Unified Video Camera, high-definition voice, ahigh-
resolution color display. Gigabit EthemeL and anew ergonomic design and user interface
that is designed for simplicitv and high usability. Accessories, sold separately, include a
color Cisco Unified IP Color Key Expansion Module and the Cisco Unified Video Camera.
Cisco Unified IPPhone 9971: This isan advanced collaborative media endpoint with
extended features, such as interactive multiparty video, high-resolution color touchscreen
display, and desktop Wi-Fi connectivity.
Cisco Unified Communications Manager Express Endpoints Implementation 3-33
Conference Stations
This subtopic explains the conference stations.
Cisco Unified IPConference Station 7936
Full-featured, hands-free conferencing
Cisco Unified IP Conference Station 7937G
G.722 wideband codec
Expanded room coverage with optional external
microphone kit
Cisco Unifiea IP Conference
Station 7936
Cisco Unified IP Conference
Station 7937G
Cisco Unified IP Conference Stations include the following models:
Cisco Unified IP Conference Station 7936: This conference station combines state-of-
the-art speakerphone conferencing technologies with award-winning Cisco voice
communication technologies. The net result is aconference room phone that offers superior
voice and microphone qualitv. with simplified wiring and administrative cost benefits. A
full-featured. IP-based, hands-free conference station, the new Cisco Unified IP Conference
Station 7936 is designed for use on desktops, conference rooms, and in executive suites.
Cisco Unified IP Conference Slation 7937: This conference station offers many
improv emcnts over the Cisco (inified IP Conference Station 7936. such as the following:
Superior wideband acoustics with thesupport ofthe G.722 wideband codec
Support for Pol-: or the Cisco Power Cube 3
Expanded room coverage ofup to 30 by 40 feet (10 by 13 meters) with the optional
external microphone kit
3-34 Implementing Cisco Voice Communications andQoS (CVOICE) v8
Identifying Cisco Unified Communications
Manager Express Endpoint Requirements
This topic describes the mechanisms that are implemented by Cisco Unified Communications
endpoints.
Endpoint Requirements
Cisco Discovery Protocol: Cisco Unified IP phones generate
and listen to Cisco Discovery Protocol messages.
DHCP: Cisco Unified IP phones can get their IPaddresses
via DHCP.
Identification by MAC address: Phonesareidentified by a
unique device ID and notbytheirIP address.
TFTP: Cisco Unified IP phones are configured automatically
by downloading device-specific configuration files from a
TFTP server.
Power: Phones can be powered over the Ethernetnetwork
cable.
Cisco Unified IP phones provide thefollowing features:
Cisco Diseovcry Protocol: Cisco Unified IP phones exchange Cisco Discovco Protocol
messages like almost all other Cisco network products. They listen to messages sent by
Cisco Catalyst switches. In this way, aCisco Catalyst switch can indirectly configure the
LAN configuration ofthe phone, including the voice VLAN and class ofservice (CoS)
settings for traffic tliat is received from an attached PC. The Cisco Discovery' Protocol
messages that are sent by the Cisco Unified IP phones are important when Cisco Unified
Video Advantage is used. Cisco Unified Video Advantage is asolution in which the phones
interactwithvideohardware and softwarethat is installed on the PC.
DHCP: Cisco Unified IPphones can have static IP configuration that isentered atthe IP
phone, oruse DHCP to obtain IP addresses that are assigned from a DHCP server.
MACaddress-based device identification: Cisco Unified IP phones areidentified bya
device ID. which is based ontheMAC address ofthe IPphone. This allows the device to
bemoved between subnets and simplifies DHCP configuration, because no specific IP
address is required for an individual phone.
TFTP: Cisco Unified IPphone configuration does not take place individually at the phone,
but isretrieved from the Cisco Unified Communications Manager Express. Cisco Unified
Communications Manager Express generates device-specific configuration files and makes
them available for download at oneor more TFTP servers. Cisco Unified IPphones will
learn theIPaddress ofthe TFTP server via DHCP, andthen loadtheappropriate
configuration file automatically aspartof their boot sequence.
CiscoUnified Communications ManagerExpress EndpointsImplementation 3-35
Power: Cisco Unified IP phones do not require wall power, but can obtain power over ihe
Ethernet from any PoE-compliant LAN switch, such as aCisco Catalyst switch. This
eliminates the need for extra power adapters and cabling on the user desk.
VC port (optional): Cisco Unified IP phones allow PCs lo be connected to the phone PC-
port and then share the uplink toward the switch. By using the voice VLAN feature of
Cisco Catalvst switches and Cisco Unified IP phones, the phone and the PC can be
separated into different VLANs onasingle access port at the LAN switch.
3-36 Implementing Cisco Voice Communications and QoS (CVOICE] v8 0 2010 Cisco Systems, Inc.
Phone Startup Process
This subtopic describes the startup process of Cisco Unified IP phones.
Phone Startup Process
Cisco IP phone obtains power from the switch.
Cisco IP phone loads locally stored image (Phone-Load).
If no local Voice VLAN ID isconfigured, theCisco IP Phone
sends out a Cisco Discovery Protocol framewith a VoIP VLAN
Query.
If the Cisco Catalyst Switch hasa Voice VLAN configured, itwill
sendouta Cisco Discovery Protocol frame with theVoice VLAN
IDfortheCisco IP phone.
Unified CME'
DHCP TFTP Server
When connected to the network, aCisco Unified IP phone goes through astandard startup
process consisting of several steps. Depending on your specific network configuration, not all
ofthese steps may occur on your Cisco IP phone:
Step 1 Obtaining power from the switeh: The Cisco Unified IP phone obtains power
from the switch, ifPoE is used. Alternatively, the IP phone can be powered by wall
power or an in-line power injector. At startup, the phone runs abootstrap loader that
loads the stored firmware image from the nonvolatile flash memory. Using this
image, the phone initializes its software and hardware.
Step 2 Loading the stored phone image: The Cisco IP phone has nonvolatile flash
memorv in which the phone firmware image is stored. At startup, the phone runs a
bootstrap loader that loads the phone image from flash memory. Using this image,
the phone initializes its software and hardware.
Step 3 Voice VLAN configuration (IP Phone): Cisco IP Phones can use 802.1Q VLAN
tagging to differentiate voice traffic from data traffic of aPC attached to the phones
PC port. The voice VLAN ID can be configured locally at the Cisco IP Phone or at
the Cisco Catalyst Switch. Ifno voice VLAN is configured locally the Cisco IP
Phone is requesting the voice VLAN ID by sending out aCisco Discovery Protocol
message that includes aVoIP VLAN Query. This message also includes the required
power for the used phone model. This allows the switch to possibly reduce the
supplied power to match the Cisco IP phones real power demand.
Step 4 Voice VLAN configuration (Switch): If avoice VLAN ID is configured on the
switch, it will respond to the received message and inform the Cisco IP Phone about
the Voice VLAN ID by also sending out aCisco Discovery Protocol message. Ifno
Voice VLAN is configured on the switch, itwill not respond with aCisco Discovery
Protocol message. In this case the IP phone will typically send out two more Cisco
Discovery Protocol messages asking fbr the Voice VLAN ID before it will continue
the boot process. Phis results in longer boot times if no voice VLAN is configured
on the switch. Ihe "sHitchport voice vian untagged" command will instruct the
su ,tch to respond uith aCisco Discovery Protocol message in order lo speed up the
phone boot process.
3-38 Implementing Cisco Voice Communications and QoS (CVOICE) v8.0 2010 Cisco Systems Inc
Phone Startup Process (Cont.)
If DHCP isenabled onthe Cisco IP Phone itreguests IP address and
TFTP server information, otherwise ituses itsstatic IPconfiguration.
v, The Cisco IP phone connects lo TFTP server and requests files in the
following order:
CTLSEP<MAC>.tlv*
SEP<MAC>.cnf.xml (SCCPconfigurationfile)
SlP<MAC>.cnf(SIP configurationfile)
~rne Certificate Tnja uaf.le s used in vwceseninty enaBleQ nuircmrnents and isnoiravereO in CVOICE vfl 0
"UniftedCME CiecoUnified ComrninitanonsManagerEipfbss
Step 5 Obtaining an IP address: If the Cisco IP phone uses DHCP to obtain an IP address,
the phone queries the DHCP server to obtain an IP address. DI1CP also informs the
IP phone about how to reach the TFTP server (DHCP Option 150). If DHCP is not
used in your network, astatic IP address and TFTP server address must be assigned
to each IP phone locally. Ifthe DHCP server does not respond, the IP phone will
make use ofthe last usedconfiguration storedin NVRAM.
Step 6 Requesting the configuration file: The Cisco IP Phone requests various files from
the TFTP server. The first file it tries todownload isthe Certificate Trust List
(CTLSEP<M-40.tlv) which is only used ifcryptographic features are enabled in
Cisco Unified Communications Manager Express.
The phone now requests its individual configuration file (SEP<jW>t0.cnf.xml)
which is only present on the TFTP server ifthe phone is already configured as an
SCCP device in Cisco Unified Communication Manager Express. Ifthis file isnot
available, it further tries todownload theSIP-bascd configuration file
(SIP<MAC>.cnfj
Phone Startup Process (Ci
If none ofthepreviously requested configuration files was found
the phonerequests a default configuration file called
XMLDefault.cnfxml
Theindividual and Ihedefault configuration file contains a
pnontized listof uptothreeUnified CM call processing nodes
and the model-specific Phone-Load-Version to use
TheCiscoIPPhonecomparesits installed Phone-Load-Version
with the Load-Version defined within thereceived configuration
file If the Load-Version is different, itrequests thenew Load-
Version fromthe TFTP server and reboots
"Unified CME Cisco Unified
CommLinicalionsManager Express
Step 7
StepS
Default configuration file: Ifthe TFTP server responds with a'"File not Found-
error message on the previously requested configuration files, the phone requests the
XMLDefault.cnf.xml file. Like the individual configuration file, also this file
contains a prioritized list of up tothree call processing nodes and the Phone-Load-
Version that is to be used foreach phone model.
Phone Load check: Once the phone received either the individual or the default
configuration file, itcompares its local Load-Version with the one specified within
the configuration file. Ifthe; aredifferent, the phone downloads the new Load from
the TFTP server and reboots.
Phone Startup Process (Cont.
The Cisco IP phone tries toregister at theUnified
Communications Manager Express call processing node.
If the Cisco IP Phone is already configured within Cisco Unified
Communications Manager Express, itwill successfully register
and will be instructed by SCCP messages to setup the display
layout (Directory Number, Softkey Buttons, Speed Dials, etc.
UrtfledCME"
Step 9 Registering on Cisco Unified Communications Manager Express: The phone
attempts to register with the highest priority call processing node on the list.
Step 10 Ifthe phone is already configured as an SCCP phone at Cisco Unified
Communications Manager Express it will successfully reg.ster and will be instructed
bv SCCP messages to set up the display layout. The display layout includes
attributes such as Directory Numbers, Softkey Buttons, Speed Dials and so on.
3-42
Step 11
Phone Startup Process {Cont
If the CiscoIPPhone is not configured inCisco Unified
Communication Manager Express, the following options are
possible1
Auto Registration isenabled - Cisco Unif edCommunications
Manager Express will dynamically create a configuration file for
the CiscoIPPhone and will request it to reboot.
Auto Registration isdisabled - Csco Unified Communications
Manager Express will not allow registration. TheCisco IPPhone
will display Registration Rejected.
"Unified CUE Cisco Unified
Communications Manager Express
If the Cisco IP Phone is nol \ct configured and received the list of call processing
nodes, from the default configuration file the following options arc possible:
Auto Registration enabled: After ihc phone tried to register at the eall processing
node. Cisco Unified Communications Manager Express, dynamically creates an
individual configuration file for this phone and requests it toreboot After reboot,
the phone will successfully register.
Auto Registration disabled: Cisco Unified Communication Manager Express will
not allow registration. The Cisco IP Phone will display a"Registration Rejected"
message on the phone displa\.
Power over Ethernet
This topic describes the three available methods to power the Cisco Unified IP phones.
Three Methods to Power Phones
Power over Ethernet (PoE):
- PoE line cards or PoE ports
- Delivers 48 V DC over data
pairs
(pins 1,2, 3, and 6) or spare
pairs (pins4, 5, 7, 8)
Midspanpower injection:
- Needs external equipment
- Delivers 48 V DC over
spare pairs
Wall power:
- Needs DC converter to
connect a Cisco Unified IP
phone to a wall outlet
Power Injedor
No Powsr _h__hh^h Pmer
Most Cisco Unified IP phone models are capable of using the following three options for
power:
PoE- With PoE. the phone plugs into the data jack that connects to the switch, and the user
PC in turn connects to the IP phone. With power-sourcing equipment (PSE), such as Cisco
Catalvst PoE-capable modular and fixed-configuration switches, power is inserted into the
Ethernet cable, such as an IP phone or IEEE 802.11 wireless access point.
Midspan power injection: Because some switches do not support PoE, amidspan power
source mav be used instead. This midspan device sits between the LAN switch and the
powered device and inserts power on the Ethernet cable to the powered device. Amajor
technical difference between the midspan and inline power mechanism is that power is
delivered on the spare pairs (pins 4.5. 7, and 8). An example of midspan PSE is aCisco
Unified IP Phone Power Injector.
Note More information about the Cisco Unified IP Phone Power Injector can be found in the
document CiscoUnified IPPhone PowerInjector at:
http//www cisco com/en/US/partner/products/ps6951/index.html.
Wall power: Wall power needs a
to a wall outlet.
DC converter for eonnecting theCiscoUnified IPphone
Note
The wall power supply must be ordered separately from the Cisco Unified IP phone.
3-43
Two PoE Technologies
This .subtopic describes the two available methods lo power the Cisco Unified IP phones over
the Ethernet network cabling.
3-44
Endpoint support
Endpoint
determi nation
Most Cisco devices (Cisco
Unifled IP phones and
wireless access points)
Uses a Cisco proprietary
FLP-based method of
determining if an attached
device requires power
Power classification !\fA
Deivery
Pcwer is deivered onlyto
devices thai require power
New range of devices
enabled by the higher
power levels
Standardizes the method
(based on voltage
measurement) of
determining if an attached
device requires power
Available as optional
element
Power is delivered only(o
deviceslhat requirepower
Cisco equipment supports the following two t>pes of inline power delivery:
Cisco original implementation of PoE: Cisco was the first to develop PoH. The original
Cisco prestandard implementation supports the following features:
Prov ides -48 VDC at up to 6.3 lo 7.7 Wper port over data pins 1. 2, 3. and 6.
Supports most Cisco dev ices (IP phones and wireless access points).
Uses aCisco proprietary method to determine if an attached device requires power.
Power is deli\ ered only todevices that require power.
802.3af PoE: Since the first deployment ofPoE, Cisco has been driving the evolution of
this technology toward standardization b> working with the IEE.E and member vendors to
create astandards-based means ofproviding power from an Ethernet switch port. The
802.3af committee has ratified this capability, "fhe 802.3af standard supports the following
features:
Specifies -48VDC at up to 15.4 Wper port over data pins 1, 2, 3. and 6orthe
spare pins 4.5. 7. and 8(aPSE can use one orthe other, but not both). Cisco
Calalj si generally provides 802.3af PoE using the data pins.
Enables anew range ofEthernet-powered devices that consume additional power.
Implementing Cisco Voice Communications andQoS(CVOICE) v8 0
- Standardizes the method of determining whether an attached device requires power.
Power is delivered only to devices that require power. This type has scyera optional
elements such as power classification, where powered devices can optionally
support asignature that defines the maximum power requirement. PSE that supports
power classification reads this signature and budgets the correct amount of power
per powered device, which will likely be significantly less than the maximum
allowed power.
Without power classification defined, the switch reserves the full 15.4 Wof power for every'
device This behavior may result in oversubscription of the available power supplies, so that
some devices will not be powered even though there is sufficient power available.
Power classification defines these five classes:
0 (default): 15.4 Wreserved
1:4 V
2:7 W
3: 15.4 W
4: Reserved for future expansion
All Cisco 802.3af-compliant switches support power classification.
The Cisco Power Calculator is an online tool that enables you to calculate the power supply
requirements for aspecific PoE configuration. The Cisco Power Calculator is available to
registered Cisco.com users at http://tools.cisco.com/cpc/LU.cpc.
Cisco Prestandard Device Detection
This subtopic illustrates how aCisco Catalyst switch detects aCisco Unified IP phone or other
inline power-capable device using ihe prestandard PoE technique.
Cisco Prestandard Device Detecti
FLP
Swtfch
It is an inline device.
Cisco Prestandard
Implementation
Powered Device P;ii:
When aswitch port that is configured for inline power detecls aconnected device the switch
sends an Ethernet East link Pulse (EI.P) to the device. The Cisco Unified IP phone loops the
II Pback tothe sw itch toindicate its inline power capability. The switch then delivers -48 V
DC PoE (inline) power tothe phone orother endpoint.
3-46 Implementing Cisco Voice Communications and QoS (CVOICE) v8
2010Gsco Syslems, Inc
IEEE 802.3af Device Detection
This subtopic describes how aCisco Catalyst 802.3af-compliant switch detects aCisco Unified
IP phone or other inline power-capable device.
IEEE 802.3af Device Detection
Switch'
uwaHfcfcE
powereddevice.
IEEE 802.3af PSE
IEEE802.3af Powered Device
The Cisco Catalyst switch detects apowered device by applying avoltage ,n the range of-2^8
Vto -10 Von the cable and then looks for a25-kohm signature resistor. Compliant powered
devices must support this resistance method. If the appropriate resistance is found, the Cisco
CataKst switch delivers power.
3-47
Cisco Catalyst Switch: Configuring PoE
This section discusses the configuration of PoE on Cisco Catalyst switches.
3-48
Native Cisco IOS Software:
CSCOIOSlconfig-if)* power inline tauto/
Use the power inline command on switches that are running native Cisco IOS Software
(examples include the Cisco Catalyst 6500. 4500, 3750, and 3560 Series Switches) The
powered device-discovery algorithm is operational in the aulo mode, fhe powered device-
discovery algorithm is disabled in the never mode. Other modes exist for allocating power
depending on the version of Cisco IOS Soflware-for example, the ability to allocate power on
a per-port basis withthe allocation milliwatt mode.
Note
The Cisco Catalyst 6500 Series Switches can run either Cisco Catalyst operating system
software or native Cisco IOS Software if the switch supervisor engine has aMultilayer
Switch Feature Card (MSFC). Otherwise, these switches can run only Cisco Catalyst
software The Cisco Catalyst 4500 and 4000 Series Switches can also run Cisco Catalyst
software or native Cisco IOS Software, depending on the supervisor engine. Generally, late-
edition supervisor engines run native Cisco IOS Software; however, the product
documentation should be checked to determine the supervisor engine and the operating
system that is supported on a specific model.
Cisco Catalyst Switch: ShowInline Power Status
The figure shows how to display the status ofinline power on aCisco Catalyst switch.
Cisco Catalyst Switch; Show Inline
Power Status
show pc
Inter
BtBthernetS/l
stEtbernetS/2
BtEthernet9/3
Oper Power ( mWatt ) Device
6300
6300
Cisco 6500 IE phone
Cisco 6500 IF Phone
n/0
Use theshow powerinline command todisplay a view ofthe power thatisallocated onCisco
CataKst switches. The switch shows the default allocated power as 10 W in addition to the
inline power status ofevery port. Thefollowing table provides a briefdescription ofthe syntax
output.
Inline Power Syntax Descriptions
show port inline power
Output Column
Description
Port Identifies the port number on the module
Inline Powered
Admin
Oper
Power Allocated
Identifies the port configuration by using the set inlinepower mod/port
[auto | off) command
Identifies ifthe inline power is operational
Detected Identifies if power is detected
mWaft
Identifies the milliwatts that are supplied on a given port
mA@42V
Identifies the milliampsat 42 Vsupplied on a given port (the actual voltage
is-48 V)
Cisco Unified Communicalions Manager Express Endpoints Implemenlation
VLAN Infrastructure
This topic describes the switch that isembedded inthe Cisco Unified IP phone and how to
implement VLANs for separation of voice fromdata traffic.
Cisco Unified IP phones have:
" Two external ports:
Switch port (P2)
PC port (P1)
* Internal circuitry to the phone module (PO)
The CiscoUnified IPphonecontains an integrated three-port 100/1000 switch, flic ports
provide dedicated connections to these devices:
Port0 is an internal 100/1000 interface that carriesthe CiscoUnified IPphonetraffic.
Port I connects to a PC or other dev ice.
Port 2 connects to the access switch or other network devices. Inline power can be obtained
at port 2,
The \ oice VLAN feature allows voice traffic from the attached IP phone and data traflic from a
daisy-chained PC to be transmitted ondifferent VLANs. Thiscapability provides flexibility and
simplicity in IP address allocation and the prioritization of Voice over Data.
If Cisco Discovery Protocol is enabled on the switch port, the switch instructs an attached
Cisco IP phone to treat the Layer 2 CoS priority value ofthe attached PC in one ofthe
following ways(basedon the extended priority that is configured at theswitchport):
Trusted: The IP phone allows the PCto send IliLL 802.3 frames (with no CoS priority
value) as well as IFlil: 802. lp frames with any CoS priority value.
Intrusted (default): The IP phone changes the CoS priority value to 0 if the PCuses
802.1p.
Configured CoS priority le\el: The IP phone sets an 802. lp header with a CoS priority
value ofx if the PC uses 802.1p with a different CoS priority level than x. or if the PC did
not use 802.1 p at all but sent 802.3 frames.
3-50 Implementing Cisco Voice Communications and QoS (CVOICE) vS.O 2010 Cisco Systems, Inc
The traflic that the IPphone sends istrusted. It can be one ofthefollowing:
802.1 Q: In thevoice VLAN. tagged with a Layer 2 CoS priority value
802.1 p: In the access VLAN. tagged with a Layer 2CoS priority value
I'ntagged: In the access VLAN. untagged with no Layer 2CoS priority value
IfCisco Discovery Protocol isenabled onthe switch port, the switch instructs the IP phone
touseone ofthe three listed options, based onthevoicevian command.
)2010CiscoSystems, Inc. CiscoUnified Communications ManagerExpress Endpoints Implementation 3-51
Voice VLAN Support
This subtopic describes voice VLANsupport.
Voice VLAN Support
A Cisco Catalyst Switch can be configured to support voice traffic in
various ways
Single VLAN access port
Multi-VLAN access port
Trunk port
Considerations
Security
Cisco IP phones/non-Cisco IP phones/IP softphones
Spanning tree
QoS
There are various methods of configuring the Cisco Catalyst switch to support voice traffic,
including the following:
Single VI.AN access port
Multi-VLAN access port
Trunk port
Various factors ha\ e to be taken into consideration, including the following:
Security
Cisco IP phones/other IP pbones/IP suftpbones (IP sotlphonc is used here as a generic term
for all software-based IP phones that are installed on a workstation.)
Spanning tree
QoS
3-52 Implementing Cisco Voice Communications and OoS (CVOICE) v8 0 2010 Cisco Systems, Inc
Single VLAN Access Port
This subtopic explains how the single VLAN access port isused toprovide voice and data
connectivity.
Single VLAN Access Port
An access port configured for one VLANonly
Typically used for IP phones or softphones other than Cisco:
IP phones other than Cisco: Use voice VLANfor access port
- Softphones Use data VLANfa access port and allow required IP
communication to voice VLAN(IP ACLs)
It used with Cisco IP phones:
Not recommended with PC attached
- If no PC attached: Use voice VLAN for access port
Voice can be lagged with IEEE 802.1 p (VLAN ID= 0) or untagged
Access Port
Untagged
Asingle VLAN accessport is the default state whenan IPphoneis connected to an
unconfigured CiscoCatalyst switch. It is typically used for IPphones otherthan Cisco, IP
softphones. or when Cisco IPphones or otherCisco voice devices donot support PCstobe
connected to them.
Whenusingthe port for sucha device,the access VLAN IDshouldbe the IDofthe voice
VLANthat is, the VLAN containing the phones. If a softphone is usedon a PC, the device
itselfthe PCcannot be in different VLANs per application {phone software versus data
applications). Therefore, the access port is usually configured for the data VLAN, and the IP
address (or subnet) ofthe PC is allowed to access VLANs with voice devices.
If a Cisco IPphonehas a PCattached, it is not recommended to put bothintothe same VLAN.
because voice and data services should be separated.
Features of a single VLAN access port include the following:
It can be configured as a secure port.
It allows physical separation of voice and data traffic using different physical ports.
It works with both Cisco and other IP phones.
The IP phone can use 802. Ip (with VLAN ID set to 0) for CoS.
Switches other than Cisco switches are typically configured in this way because they do not
usually support the voice VLAN feature.
) 2010 Cisco Systems, Inc. Cisco Unified Communications Manager Express Endpoints Implementation
Multi-VLAN Access Port
This subtopic describes a multi-VLANaccess port and howyou can use il to connect lo an 11
phone.
lulti-VLAN Access Port
An access port able to handle two VLANs
Access (data) VLANand voice (auxiliary) VLAN
Voice traffic is tagged with IEEE 802 10 VLAN ID:
Data traffic is untagged and is forwarded by IP to and from PC port.
Phone can be hardened to prevent PC from seeing the voice traffic (by
default, phone acts like a hub).
Best choice with Cisco IP phones.
Voice VLANdoes not need to be configured on IP phone, but it can be
learned from Cisco Discovery Protocol messages senl out by Ihe switch.
Access Port
Tagged 802.1Q
Untagged
All Cisco Catalvst switches support multi-VLAN access ports. All data devices typically reside
on data VLANs in the traditional switched scenario. Aseparate voice VLANmay be needed
when combining the voice network into the data network. While on Cisco IOS Software-based
Catalyst switches, the voice VLAN is also called voice VLAN in configurations, Cisco Catalyst
switches that are using the Catalvst operating system refer to the voice VLAN as the auxiliary
VLAN. The new voice VLAN can be used to represent Cisco IP phones. Although it is a voice
VLAN. in the future, other types of nondala devices will reside in the voice VLAN.
The placement of nondala devices, such as IP phones, in a voice VLAN makes it easier for
customers to automate the process of deploy ing IP phones. IP phones will boot and reside in the
voice VLAN if the switch is configured lo support them, just as data devices boot and reside in
the access (data) VLAN. The IP phone communicates with the switch via Cisco Discovery
Protocol when it powers up. The switch provides the IP phone with the appropriate VLAN ID,
You can implement multiple VLANs on the same port by configuring an access port. A lagging
mechanism distinguishes among VLANs on the same port. 802.IQ is the IFTT standard for
tagging frames with a VLAN ID number. The IP phone sends tagged 802.10 frames. The PC
sends untagged frames, and the switch puts the frame into the configured access VLAN. When
the switch receives a frame from the network that is destined for the PC. it removes the access
VLAN tag before forwarding the untagged frame to the PC.
The following are some advantages of implementing dual VLANs:
A multi-VLAN access port can be configured as a secure port.
A voice VLAN ID is discovered using Cisco Discovery Protocol, or it is configured on the
IP phone.
3-54 Implementing Cisco Voice Communications and OoS (CVOICE) vS.O 2010 Cisco Systems Inc
Dual VLANs allow for the scalability ofthenetwork, from anaddressing perspective. IP
subnets usually have more than 50 percent (often more than 80 percent) oftheir IP
addresses allocated. Aseparate VLAN (separate IPsubnet) tocarry the voice traffic allows
the introduction of many new devices, such asIPphones, into the network without
extensive modifications to the IP addressing scheme.
Dual VLANs allow for the logical separation of dataandvoice traffic, which allows the
network to processthese twotraffictypes individually.
Implementing dual VLANs allows youtoconnect two devices thatarcindifferent VLANs
to a single switch port.
12010 Cisco Systems, Inc. Cisco Unified Communications Manager Express Endpoints Implementation 3-55
Trunk Port
This subtopic describes a trunk port andhow youcanuse il loconnect to an IPphone.
Trunk Port
Atrunk port is able lo handle multiple VLANs
Usually, data traffic is untagged and put into a native VLAN.
Data traffic can be tagged wilh any 802.1Q VLAN IDif supported by PC (and
permitted by IP phone).
Voice VLAN does not need to be configured on IP phone, but it can be learned
from Cisco Discovery Protocol messages sent out by the switch.
Secunty considerations:
Cannot be confgured as secure port.
If allowed VLANs are not limited. PC has access to all VLANs of the switch.
Trunk Port
Tagged 802.10
Untagged (Native VLAN)
Rather than a dual VLAN access port, you can use a trunk port for connecting a switch to an IP
phone. Becausea Cisco Catalyst switch supports multi-VLANaccess ports, a trunk port is not
commonly used to connect a switch to a Cisco IP phone. However, a trunk port can also be a
way to connect a Cisco 1P phone to a switch other than Cisco. Some of ihe first Cisco switches
supported voice VLAN features, allowing the voice VLAN ID lo be used by a phone via Cisco
Discovery Protocol only on trunk ports.
When an 802. IQ trunk port is used, frames ofthe native VLAN are always transmitted
untagged and should be received untagged. In other words, a PCwhich usually does not send
802.IQ frames but rather untagged Ethernet framesis part ofthe native VLAN. while the
Cisco IP phone tags its frames with 802.IQ. However, a PC could send and receive tagged
frames and thus access all VLANs that are configured in the switch.
On trunk ports, tagged frames arc permitted by default. Therefore, the only function of this
command is to allow the IP phone to learn the VLAN ID that should be used for its traffic by
Cisco Discovery Protocol (although nol required because it can be manually configured at the
phone). Some ofthe considerations when implementing a trunk port to support Cisco IP phones
are as follows:
On some end of life (LOL) Cisco IOS and Catalyst switches, Portl'ast cannot be enabled on
a trunk port.
The port cannot be configured as a secure port.
fhe PC can access all VLANs if it supports 802.IQ.
3-56 Implementing Cisco Voice Communications and QoS (CVOICE) v8.0 2010 Cisco Systems. Inc
Ethernet Frame Types Generated by Cisco IP Phones
This subtopic describes the different Ethernet frame types that aCisco IP phone generates
basedon the switch port configuration.
Ethernet Frame Types Generated by
Cisco IP Phones
Single VLAN
Access Pan
Destination MAC
6B
SourcftUAC
88
Ethertype
2B
Payload
46-1600B
D est nation MAC
6B
Mull-VLAN
Access Pol
O est nation MAC
6B
Single VLAN
Access Pert with
802.1p Configuration
SotmaMAC
SB
B02 1Q
4B
Bhortypo
2S
Payload
4&-1500B
TPID
16 b 3b iblHi
016100 0-7 0,1 |l40|
SourcaMAC
6B
BQ2 IQ
4B
Elhertypt
2B
Payload
46-1500B
TPID
16 b
SQZIp
36 ^HHI
0S100 0-7 01I 0 |
TPID- Tafl Protocol Idtmfpti
CFI Cin&nicil Formn 1I denbller
FCS- FramBCriecKSMiHnce
Based ontheswitch portconfiguration that is used toconnect a CiscoIPphone, the following
Ethernet frame types are present:
Single VLANaccess port: If the switchport is configured as a singleVLAN accessport
only, standardEthernet V2 frameswill be generated by the Cisco IPphoneand the Cisco
Catalyst switch for voice traffic. There is no VLAN ID nor CoS information present within
the transmitted frames. CoS classification can be configured on the switch.
Multi-VLAN access port and trunk port: Both port types will cause the Cisco IP phone
andthe CiscoCatalystswitchto generate standard-based 802,IQ frames to tag voice
VLAN trafficaccordingly. Because 802.IQ includes 802.1p,CoSmarkings arc included in
these frame types.
If the switch port is configured as a multi-VLANaccess port, only voice VLAN-tagged
frames and untagged frames (native VLAN for data traffic) are present. In a trunk port
configuration, tagged frames forotherVLANs tliat might beconfigured on theswitch will
also be sent out on the switch port. This can be prevented by specifying allowed VLANs.
Single VLAN' access port with 802.1p configuration: In a single VI ANaccess port with
additional 802.lp CoSconfiguration, standard802.IQ framingwill be used. The difference
between the framing of a multi-VLANaccess port or a trunk port and a single VLAN
access port with802.lp configuration is that the latter will alwaysuse 0 for the VLAN ID.
i2010 Cisco Syslems, Inc Cisco Unified Communicalions Manager Express Endpoints Implementation 3-57
Blocking PC VLAN Access at IP Phones
Thissubtopic describes the blocking PCaccessat IPphones
With default configuration on a trunk port, if PC sends 802.1Q
tagged frames, all VLANs can be accessed from PC:
Disable voice VLANaccess al phone.
Prevent PC from sending and receiving dala lagged with voice VLAN ID.
Other VLAN IDs are permitted on some IP phones.
Disable span to PC port (on supported IP phones).
Preven! PC from sending and receiving any 802.IQ tagged frames.
Tagged 802 1Q
{Voice VLAN 10)
Untagged (Access VLAN 20)
Whena PC is connected to an IP phone, there are two potential security issues:
If the switch port is configured as a trunk, the PC has access to all VLANs.
If the switch port is configured as an access port, fhe VC has access to the voice VLAN.
The reason for this is that, by default, the IP phone forwards all frames that are received
from the sw itch to the PC and \ ice versa.
You can configure Cisco IP phones to block access by the PC to the voice VI AN. If
configured, the IP phone will not fonvard frames that are lagged with the voice VLANID. This
configuration solves PCVLANaccess issues with dual VLANaccess ports, because the PCis
limited to using the access VLAN (untagged frames).
Limiting VLANs on Trunk Ports at the Switch
This subtopic describes limitingVLANs on trunk ports.
Limiting VLANs on Trunk Ports at the Switch
VLANs allowed on a trunk port can be configured at the switch.
Recommendation is to only allow native VLAN and voice VLAN:
- It blocks PC access to all other VLANs, independent of IP
phone configuration and model.
- Access to voice VLAN can only be prevented by IP phone
configuration but is supported on all IPphone models with PC
ports.
- It improves performance.
Trunkportson CiscoCatalystswitchesshouldbe configured to allowonlythe necessary
VLANs. In a CiscoIPphonewithan attachedPC, theseVLANsare the voice VLAN and the
native VLAN. Denying all other VLANs provides the following advantages:
Increased security: It is a best practice to only allowthose VLANs on a switch port that
are used by the connectingend devices. Access to voice VLAN can only be prevented by
IP phone configurationbut is supported on all IP phone models with PCports.
Increased performance: Reducingthe number of VLANs cuts down unnecessary
broadcast traffic.
Increased stability: Limiting the numberof VLANs will also minimize potential Spanning
Tree Protocol (STP) issues and increase network stability.
Configuring Voice VLAN in Access Ports Using Cisco IOS
Software
This topic shows how to configure voice VI AN in access ports on Cisco Catalyst switches
using native Cisco IOS Software.
Configuring Voice VLANs in Access
Port Using Cisco IOS Software
Example 1 (single VLAN access port):
Console(coofig)tinterface FastEthemet0/1
Console(config-if)#switchport mode access
Console Iconfig-if1#swi tchport voice vlandotlp
Console(config-if|#switchport access vian 261
Example 2 (multi-VLAN access port):
Console (conf ig] (finterf ace FastEthernetO/1
Console(config-if)#switchport mode access
Console(config-if)ffswitchport voice vian 261
Console(config-if)Sswitchport access vian 262
L'se the commands that are shown in the figure to configure voice and data VLANs on the
single port interface of a switch that is running native Cisco IOS Software.
The first example shows the configuration of a single VLAN access port, fhe switch is
configured to transmit Cisco Discovery Protocol packets to enable the Cisco IPphone to
transmit voice traffic in 802.lp frames that are tagged with VLAN ID 0 and a Layer 2 CoS
value, flic switch puts the 802. lp voice traffic into the configured access VLAN. VLAN 261,
which is used for voice traffic.
The second example shows a multi-VLAN access port configuration in which the voice traffic
is sent to VLAN 261 and the data is using the access VLAN 262.
Note The multi-VLANaccess port is the recommended configuration for Cisco IF phones that
have a PC port
Catalyst Switch Voice Interface Commands
Command Description
switchport mode access
Configuresthe switchport to be an access (nontrunking) port.
spanning-tree portfast
Causes a port to enter the spanning-tree forwarding state
immediately, bypassing the listening and learning states. You
can use Portfast on switch ports that are connected to a
single workstation or server (as opposed to another switch or
network device) to allow those devices to connect to the
network immediately.
awitchport access vian
data VLAN ID
Configure the interface as a static access port with the access
VLANID(262 in this example); the range is 1 to 4094.
switchport voice vian
{voice_vlan_ID I dotlp |
none untagged}
When configuring the way in which the Cisco IP phone
transmits voice traffic, note the following syntax information:
Enter a voice VLAN ID to send Cisco Discovery Protocol
v2 packets that configure the Cisco IP phone to transmit
voice traffic in 802.1Q frames that are tagged with the
voice VLANIDand a Layer 2 CoS value {the default is 5).
Valid VLAN IDs are from 1 to 4094. The switch puts the
802.1Q voice traffic into the voice VLAN.
Enter the dotlp keyword to send Cisco Discovery
Protocol v2 packets that configure the Cisco IP phone to
transmit voice traffic in 802.1 p frames that are tagged with
VLAN ID0 and a Layer 2 CoS value (the default is 5 for
voice traffic and 3 for voice control traffic). The switch puts
the 802.1p voice traffic into the access VLAN.
Enter the untagged keyword to send Cisco Discovery
Protocol v2 packets that configure the Cisco IP phone to
transmit untagged voice traffic. The switch puts the
untagged voice trafficinto the access VLAN
Enter the none keyword to allow the Cisco IP phone to
use its own configuration and transmit untagged voice
traffic. The switch puts the untagged voice traffic into the
access VLAN.
) 2010 Cisco Systems, Inc. Cisco Unified Communications Manager Express EndpoinIs Implementation 3-61
Configuring Trunk Ports Using Cisco IOS Software
This topic shows how to eoniigure trunk ports on Cisco Catalyst switches using native Cisco
IOS Software.
Configuring Trunk Ports Usini
Cisco IOS Software
Example 3 (trunk port):
Console (config) ((Interface FastEthernetQ/1
Console(config-ifI (tswitchport trunk encapsulation dotlq
Console(config-ifI(tswitchport mode trunk
Console (config-if) (tswitchport trunk native vian 262
Console (config-if) (tswitchport voice vian 261
Console(config-if) (tswitchport trunk allowed vlan261,262
L'se the commands that are shown in the figure to eoniigure the trunk interface of'a switch that
is running native Cisco IOS Software.
In the example. VLAN 261 is used for voice tratlle: VLAN 262, which is also the native
VLAN. is used for data traffic. All other VLANs are blocked from the trunk interface.
Note The native VLANdoes not have to be permitted in the allowed VLANlist
Catalyst Switch Voice Interface Commands
Command Description
switchport mode trunk
Configures the switch port to be a trunk port.
switchport trunk
encapsulation dotlq
Configuresthe switchport trunkencapsulation to 802 1Q
instead of leaving it as autodetect.
switchport trunk native
vian VLAN-ID
Configures the interface native VLAN. When you use an
802.1Q trunk port, all frames are tagged except those on the
VLAN that are configured as the native VLAN for the port.
Frames on the native VLANare always transmitted untagged
and are normally received untagged.
trunk
Causes a port to enter the spanning-tree forwarding state
immediately, bypassing the listening and learning states. You
can use the portfast command on switch ports that are
connected to a single workstation or server (as opposed to
another switch or network device) to allow those devices to
connect to the network immediately.
switchport trunk allowed
vian VLAN-ID
Specifies the VLANs that are allowed on the trunk port.
12010 Cisco Systems. Inc. Cisco Unified Communications Manager Express Endpoints Implementation
Verifying Voice VLAN Configuration Using Cisco IOS Software
This subtopic describes howto verify voice VLANconfigurationon Cisco Catalvst switches
that use nati\ e Cisco IOS Software.
Verifying Voice VLAN Configurate
Usinq Cisco IOS Software
Claas-l-Switchtshow interfaces faO/4 switchport
Name: Fa 0/4
Switchport: Enabled
Administrative Mode: static access
Operational Mode: static access
Administrative Trunking Encapsulation: negotiate
Operational Trunking Encapsulation: native
Negotiation of Trunking: Off
Access Mode VLAN: 262 (VLAN02S2)
Trunking Native Mode vlwj; l (default)
Voice VLAN: 261 (VLAN0261)
You can verilv \ oice VLAN configuration on Cisco Catalvst switches that are running native
Cisco IOS Software b> using the show interface mod'port switchport command.
The figure shows that interface fa()'4 is configured as an access port with access VLAN 262
and voice VLAN 261. Also, this port is using the default native VLAN ID I.
Implementing Cisco Voice Communications and QoS (CVOICE| v8 0 2010 Cisco Systems, Inc
IP Addressing and DHCP
This topicdescribes howto assignaddresses to Cisco UnifiedIPphones.
Voice Segment Addressing
Separate IP prefix
recommended for
manageability and
security reasons
Cisco Unified IP
Phone uses separate
logical network.
CiscoUnified IP Phone and PC
areon the same physical switch
port.
Cisco Unified IP phones require network IP addresses. The IP addresses assigned to the phones
should be assigned from separate subnets for easier manageability and security. In most
scenarios, the following guidelines should be followed when deploying IP addresses:
Fxisting IP address subnets should be used for data devices (PCs, workstations, servers).
DHCP should be used to assign IP addresses to Cisco Unified IP phones.
Separate IP subnets should be used for phones, if available in the existing address space.
Private address space (defined in RFC 1918), such as the 10.0.0.0 network, can be used for
the voice VLAN if other subnets are not available.
) 2010 Cisco Systems, Inc. Cisco Unified Communications Manager Express EndpoinIs Implementalion
DHCP Parameters
This subtopic explains the required DHCP scope parameters assigned lo Cisco Unified IP
phones.
DHCP Parameters
ADHCP scope can be configured on the Cisco
Unified Communications Manager Express router or
any other DHCP platform. The scope should define
the following:
Range of available IP addresses
Subnet mask
Default gateway
Addresses of the TFTP servers (option 150)
DNS servers, if using naming instead of IP addresses
When DHCP is used to dynamicall) assign IP parameters lo the Cisco Unified IP phone, the
DHCP server can be implemented either on the Cisco Unified Communications Manager
Express, another Cisco IOS router, or any DHCPserver in the network. The DHCPscope must
include a range of IP addresses with the subnet mask, the default gateway, and the addresses of
the TFTP servers, which arc carried using option 150. Optionally, the DHCP scope can also
specif) the DNS server addresses.
The following messages are involved in a DHCP exchange:
DIICPD1SCOVRR: By default, the Cisco Unified IP phone (DHCP client) sends a
DHCPDISCOVFR request to the 255.255.255.255 broadcast address on the acquired voice
VLAN.
DIKTOITKR: The server assigns a free IP address with the remaining required
parameters for the scope. The offer is sent to the DHCP client (the phone) using the
broadcast address 255.255.255.255.
DHCP selttngs initialized: The phone takes the values, received Iromthe DHCPresponse
and applies themto the IP stack ofthe IPphone, and then sends a Gratuitous ARP lo
normalize the ARP cache for other dev ices on the network.
Configuration requested from TTTP server: The phone uses the value, received in
option 150. to retrievea configuration file from the TFTPserver. The CiscoUnified
Communications Manager Express router is typically the TFTP server, although an external
TFTP server can also be used instead ofthe Cisco Unified Communications Manager
Express svsleni.
Router Configuration with 802.1Q Trunk
This subtopic describes how toconfigure a Cisco IOS router for DHCP and 802.1Q trunking
necessary to support voice and data VLANs.
Router Configuration with 802,1Q Trunk
802 1Q Trunk
(VU\N 11,111)
Tagged 802 IQ (Voice VLAN111
Untagged 802.3 (Native VLAN11)
lp dhep aicludad-addteaa 10.11.0.1 10.11. 0.10
ip dbcp aicludad-addi-aaa 10.111.0.1 10.111,0,10
lp dhep pool Data
n.tvotk 10.11.0.0 355-355.355.0
d.f.ult-rout.r 10 11.0.1
dna-aarvar 10.9.9 8 10-9.9-9
lp dhp pool Phonaa
natvork 10.111.0. 255.355.255.0
dafault-routar 10 111.0.1
option ISO lp 10. 11.0.1
dna-aarvar 10.9.9 S 10.9.9.9
lntarfaca FaatEttiarB to/o
no lp Bddraaa
inUifui Put Ithaca.t0/0.11
ausapiulaCisii dotlq 11
lp addraaa 10.11.0. 155.355.355.0
intarfaca Paatltharii .t0/0.111
ancapaulation dotlq 111
lp addraaa 10.111.0 1 355-255.255.0
The figure illustrates a Cisco IOSrouter that acts as DHCP server and has an interface that is
configured for 802.IQ trunking necessary to support voice and data VLANs. The DHCP server
is configured with two scopes: for the phone subnet (Phones) and for the PC subnet (Data). The
phone scope uses the command option 150 ip 10.111.0.1 to indicate the TFTP server address.
In this example, the TFTP server address is a local interface address, which is common for
Cisco Unified Communications Manager Express deployments. DHCP, TFTP, and Cisco
Unified Communications Manager Express could run on the same Cisco IOS router, or be
distributed. The router has two interfaces that correspond to the 802. IQ tags ofthe voice and
data VLANs.
)2010 Cisco Systems, Inc. Cisco Unified Communications Manager Express Endpoints Implementation 3-67
Router Configuration with Cisco EtherSwitch Network Module
This subtopic explains how to configure the router with a Cisco EtherSwitch network module
installed.
Router Configuration with Ci
Tagged802 1Q (Voice VLAN 111)
Untagged802 3 (Native VLAN 11)
interface FastEthernetl/1
switchport access vian 11
switchport voice vian 111
interface vlanll
ip address 10.11.0.1 255.255.255.0
I
interface Vlanlll
ip address 10.111.0.1 255.255.255.1
DHCP configjraiion part is omitted
The figure shows a router with an installedCisco LtherSwilchmodule. With integrated switch
components on the router. I,a\er 3 interfaces are defined using the interface vian command.
The VLANs are then applied to the physical ports using the switchport command, because
they are used on Cisco IOS switches. The DHCPconfiguration is omitted because it is identical
to the previous example.
DHCP Relay Configuration
This subtopic explains the DHCP relay feature and how it is used inthe Cisco Unified
Communications system.
DHCP-Retay Configuration
iotarfsoa FastltbernetO/O
no ip address
interface FastEtoarnet 0/0.11
encapsulation dotlq 11
ip addrasa 10.11.0.1 255.255.255.0
ip halpsr-addrasa 10.1.1.1
interface FastEtharnet 0/0.111
encapsulation dotlq 111
ip address 10.111.0.1 255.255.255.0
ip bslpar-sddran 10.1.1.1
The DIICPrelay agent is a devicethat relays DHCP messagesbetweenclientsand serverson
different IP networks. It is a Cisco IOS router that "listens" to DHCP client messages being
broadcast on the subnetand relays themto the configured DHCP server. The DHCPserver then
sendsresponses usingDHCPrelay agent backto the DIICPclient. The DHCP relay agent
saves the administrator the effort of installing and running a DHCP server on each subnet.
Thefigure shows a Cisco IOSrouter thathas thevoice anddataVLANs directly attached to it.
It acts as the DIICPrelay agent for the voice and data subnets and has the ip helper-address
command that is configured on the respective voice and data interfaces. The ip helper-address
pointsto the DIICPserver and is necessary to convert the DHCPbroadcasts to unicasts sent to
the DHCPserver. The DHCPserver has pools that are configured for two subnets (voice and
data) that arc not directly connected to it.
2010 Cisco Systems, Inc. Cisco Unified Communicalions Manager Express Endpoints Implementation 3-69
Network Time Protocol
This topic describes the Network Time Protocol (NTP) and its significance in the Cisco Unified
Communications s\ stems.
The system time of the Cisco Unified Communications
Manager Express should be synchronized with the entire
Cisco Unified Communications network:
Accuracy required for
Phone display: Time received from the Cisco Unified
- Call lists (missed/received): Time ofthe call
Voice mail: Time when the message was left
Reporting and troubleshooting Time stamps for syslog and
traces
Call Detail Records Billing
NTP
Widespread use RFC standard
NTP hierarchy NTP server can have STRATM clock or
synchronize with a more authoritative source
The internal dock of a Cisco IOS routercan drift
The time clock should be synchronized in all components ofthe Cisco Unified
Communications network. Time accuracy is needed for a number of aspects, such as the
following:
Phone display: Cisco Unified IP phones display the lime as it is received from the Cisco
Unified Communications Manager Lxpress.
("all lists: Cisco Unified IP phones list the missed, received, and placed calls, including the
time that the eall occurred.
Voice-mail: Voice-mail s\ stems provide the time when the message was left.
Reporting and troubleshooting: Reporting data is typically collected on central systems
that order the infonnation based on the time that an event is received. Such data is marked
with time stamps for future use. fhe time stamps must be reliable and accurate for effective
troubleshooting and monitoring.
Billing: Call Detail Records (CDRs) are used lo report infomialion about the calls. This
data can be sent to billing applications.
NTP is a widespread Internet Fngineering Task Force (IF.TF) standard that supports a hierarchy
of clock sources that \ar\ in the le\el of trust. Trusted servers are typically highly available
systems that arc equipped with extremely reliable clocks, such as atomic sources. NTP is
strong!} recommended to be used instead ofthe internal router clock, which can drift and does
not allow a precise synchronization with other network components. NTP synchronizes the
Cisco Unified Communications Manager bxpress router to a single clock on the network,
known as the master clock.
NTP Configuration Example
This subtopic describes how toconfigure NTP on a Cisco IOS router.
NTP Configuration Example
Cisco Unif ed IP phone lime comes
from Ihe Cisco Unified
Cisco Unified Communicalions
Manager Express receives time
from the NTP seivers.
101 21
10.1 1.1
10 2.2.2
clock timezone PST -8
clock summer-time lone PST reourring first Sunday narch 02:00
last Sunday October 03:00
ntp server 10.1.1.1 prefer
ntp Burver 10.2.2.2
Theexample inthe figure shows the Cisco Unified Communications Manager Hxpress router in
the Pacific Standard time zone withdaylightsavingtime turnedon. The router is set to
synchronize itssystem time to the external time server 10.1.1.1 and 10.2.2.2, while theformer
server is the preferred NTP source.
) 2010 Cisco Systems. Inc. Cisco Unified Communications Manager Express Endpoints Implementation 3-71
Endpoint Firmware and Configuration
fhis topic describes how the Cisco Unified IPphones obtain their configuration and firmware
image.
Phone requests its specif c configuration using its MAC ID.
Response with MAC-specific configuration file
No response- fallback lo previous configuration
Response with no MAC-specifc configuration file
Phone requests the default configuration
Name of phone firmware may be included in configuration file (either specific
or default)
-^Lj TFTP request for CTLSEP-=MAC> Hv ^
TFTP requesl lor SEP* MAC? cn( xml (SCCP configuration Me) _
TFTP requesl for SIP^MAO.cnf (SIP configuration file)
Obtaining
configuration
tile
Phone ctiec-s ^
cunent firmware "
and downloads
if different
A response with file, B no response. C. response with no file
ifC TFTPrequestforXMLDefauHcnf xml
ifC TFTPre5ponsew(inXMLDelaultcnf xml
TFTP request lor firmware
TFTP response with firmware
The TFTP server has device-specific and generic configuration files. A configuration file
includes parameters for connecting to Cisco Unified Communications Manager Express and
information about which image load a phone should be running.
First, the phone requests the CTI.SF.P<MAC>.tlv that contains a certificate trust list not
covered in this course. Then it requests its MAC-address-specific SCCP/SIP configuration file:
first SFP<niac>.cnf.\ml and then SIP<MAC>.cnf. If the TFTP server does not respond, the IP
phone falls back to the last used configuration stored in NVRAM. If the phone is new, this file
will not be found, because the phone is not currently configured in the Cisco Unified
Communications Manager Express database. In that case, the TFTP server responds without
pro\ iding the device-specific configuration file. The phone then requests the generic
XMLDefault.enf.xml file.
The phone requests the .loads file, if one was specified in the configuration tile (specific or
default), to see what image the phone should be running. IIThe .loads file specifies an image
that is different from the image thai is stored in the phone NVRAM, the phone attempts to
obtain the new image from the TFTP server. If the image is downloaded and verified
successfully, the phone reboots to load the new image and then to register to the priman' Cisco
Unified Communications Fxpress svstem.
Downloading Firmware
This subtopic explains how toconfigure the TFTP service for phone image files.
Downloading Firmware
router(config)#
Itftp-server location:filename
Allows a tile to be downloaded using TFTP
tftp-server lash:appB4S .9-0-2ES2.sba
t ftp- server Clash:cnu45 . 9-0-2E32 .aba
tftp-server lash:cvm45sip.9-0-2E32.Bba
tftp- server f lash;dsp45. 9-0-2ES2 . sbn
t ftp-server flash:! ar45sip. 9-0-2ES2 . sbn
tftp-server flash: SIP45. 9-0- 29B1S. loads
tftp-server flash: term45.default.loads
tftp- server flash: tennSS .default .loads
The tftp-sener location: filename commandallows the file, specifiedusingthe location and
filename parameters, to bedownloaded usingTFTP. ForCiscoUnified Communications
Manager Express, youmustconfigure the finnware files that will be downloaded byendpoints
to be available through TFTP.
The example shows howto configure the TFTP service for the files belonging to the SIP
firmware package of Cisco Unified IP Phones 7945 and 7965 version 9.0.
Firmware Images
This subtopic explains how finnwareimage files are packaged.
Firmware images are implemented as file bundles.
The firmware package includes .loads loader file that describes
the components ofthe bundle.
Package examples:
cmterm-7945_7965-sip.9-0-2SR1 (SIR 7945/65, v9.0)
- cmterm-7945_7965-sccp.9-0-2SR1 (SCCP, 7945/65, v9.0)
finnware images arc implemented as file bundles that contain multiple images for various
components ofthe Cisco Unified IP phone.
The finnware package includes the .loads loader file that describes the components ofthe
bundle. This file is downloaded by the phone first and it tells the phone which files should be
requested from the TFTP sener. The phone learns the name ofthe appropriate .loads file from
the configuration file. The configuration file obtains the information from the configured load
command. The load command references the appropriate .loads file.
Following are a few examples of firmware packages:
cmterm-7945_7965-sip.9-0-2SRI (SIP. 7945/65. v9.0). This package includes these
individual files: apps45.9-0-2FS2.sbn. cnu45.9-0-21:.S2.sbn, evm45sip.9-0-2FS2.sbn,
dsp45.9-0-2FS2.sbn.jar45sip.9-0-2FS2.sbn. SlP45.9-0-2SRIS.!oads. tcrm45.dcfault,loads.
temi65. default, loads.
cmtcrm-7945_7965-sccp.9-0-2SRl (SCCP. 7945/65. v9.0).
Setting up Cisco Unified Communications
Manager Express in SCCP Environment
This topic describes howto configure the Cisco Unified Communications Manager Express to
suppon SCCP endpoints.
Configuring IP Protocol Mode
router(config)#
telephony-service
Enters telephony service (Cisco Unified Communications
Manager Express for SCCP endpoints) configuration mode
router(eonfig-telephony)#
protocol mode {ipv4 ipvfi dual-stack [preference
{ipv4|ipvS}]}
Sets the operational mode for phones
Default: IPv4
Setting up the Cisco Unified Communications Manager Express system manually involves
using the command-line interface (CLI). This type of setup enables you to leverage existing
knowledge of Cisco IOS Software to implement Cisco Unified Communicalions Manager
Express functions. You can view, back up, and restore the configuration using a simple text
file. Manual setup can save time and effort when you use it for multiple site deployments
because you change only the settings that are different on each site.
The telephony-service command enters the configuration mode for systemwide parameters for
SCCP IP phones in Cisco Unified Communications Manager Express.
The protocol mode command is used to configure SCCP phones in IPv4-only, IPv6-onIy. or
dual-stack mode. For dual-stack mode, the user can configure the preferred family, IPv4, or
IPv6. For a specific mode, the user is free to configure any address and the system will not hide
or restrict any commands on the router. On a per-call basis and based on the configured mode,
SCCP phones choose the right address for communication.
For example, if the DNS reply has both IPv4 and IPv6 addresses and the configured mode is
IPv6-onK (or lPv4-only). the system discards all IPv4 (or IPv6) addresses and tries the IPv6 (or
IPv4) addresses in the order that they were received in the DNS reply. If the configured mode is
dual-stack, the system first tries the addresses ofthe prefened family in the order that they were
received in the DNS reply. If all ofthe addresses fail, the system tries addresses ofthe other
familv.
i 2010 Cisco Systems, Inc. Cisco UnifiedCommunicalions Manager Express Endpoints Implementation
Configuring Source IP Address and Firmware Association
This subtopic explains howto configure the source IPaddress and the SCCP endpoint finnware
association.
3-76
Configuring Source
j Associate
router(config-telephony)#
ip source-address {ipv4 address ; <ipv<i address} port]
[secondary {ipv4 address \ ipvS address} [rehome seconds]]
[any-match j strict-match]
Configures the source IP address of the Cisco Unified Communications
Manager Express (mandatory)
Port option defines Ihe SCCP port, default is 2000
* Source address may be the same as TFTP server
Secondary option defines the backup Cisco Unified Communications
Manager Express
- strict-match accepts phone registrations targeted at Ihis address only
router (conf lg-telephony) #
load model firmware-file
Optional, associates a phone type wilh a firmware file (without suffix)
Adds firmware specification lo configuration file
For example load 7965 SCCP45 9-0-2SR1S
The ip source-address command enables a router lo receive messages from Cisco Unified IP
phones through the specified IP address and port. The Cisco Unified Communications Manager
Fxpress router cannot communicate with Cisco Unified Communications Manager Express
phones if the IP address ofthe port to which they are attached is not configured. The Cisco
Unified Communications Manager F.xpress router can receive messages from IPv6-enabled or
IPv4-enabled IP phones or from phones in dual-stack {both IPv6 and IPv4) mode. Thedefault
port is 2000. The configured IP address may or may not be the same as Ihe TFTP server
address. The secondary option allows the configuration ol'lhe second Cisco Unified
Communications Manager Express router with which phones can register if the primary Cisco
Unified Communications Manager F.xpress roulcr fails. The strict-match ke\word instructs the
router to reject IP phone registration attempts if the IP server address used by the phone does
not match the source address.
The load command updates the Cisco Unified Communications Manager Express configuration
file for the specified t\pe of Cisco Unified IP phone to add the name ofthe firmware file to be
loaded by a particular phone type. The finnware filename also provides the version number for
the phone finnware that is in the file. A separate load command is needed for each type of
phone.
The following list shows the supported phone models for which you use the load command:
Note Do not use the file suffix when using the load command.
7902: Cisco Unified IP Phone 7902G model
7905: Cisco Unified IP Phone 7905(5 model
7906: Cisco Unified IP Phone 7906CJ model
Implementing Cisco Voice Communications and QoS (CVOICE] v8 0 2010 Cisco Systems Inc
7910: Cisco Unified IP Phone 79I0G+SW model
7911: Cisco Unified IP Phone 791 lGmodel
7912: Cisco Unified IP Phone 79I2G model
7914: Cisco Unified IP Phone 7914 Expansion Module
7920: Cisco Unified Wireless IP Phone 7920 model
7921: Cisco Unified Wireless IP Phone 7921G model
7935: Cisco Unified IP Conference Station 7935 model
7936: Cisco Unified IP Conference Station 7936 model
7960-7940: Cisco Unified 7960G and 7940G models
7941GE: Cisco Unified IP Phone 7941G-GE model
7961GE: Cisco Unified IP Phone 7961G-GE model
7971G-GE: Cisco Unified IP Phone 7971G-GE model
7985: Cisco Unified Video IP Phone 7985G model
ATA: Cisco ATA 186 and 188 analog telephone adapters
2010 Cisco Systems, Inc. Cisco Unified Communicalions Manager Express EndpoinIs Implementation 3-77
Enabling SCCP Endpoints
This subtopic explains how to enablethe Cisco Unified Communications Manager Fxpress fbr
SCCP endpoints.
Enabling SCCP Endpoints
max-ephones maximum-ephoneb
Sets the maximum number of ephones that may be defined in the system
For efficient use of memory, should not be higher than needed
Default 0
router(conflg-telephony)#
I max-dn
Sets the maximum number of ephone-dns that may be
defined in the system
For efficient use of memory, should not be higher than
needed
Default. 0
An SCCP endpoint is defined within the Cisco Unified Communications Manager Express as
an ephone. The maximum number of ephones depends on the hardware platform ofthe Cisco
Unified Communications Manager Express. The max-ephones command is set to 0 by default
to conserve s\stem memory. If you set this value above the required number of directory
numbers, the router reserves system memory that it could use for other functions. Use the max-
ephone ? command in telephony-service configuration mode to determine the maximum
number of ephones supported b_\ the hardware. Set the value within the range that complies
with the license.
Before am directory numbers can be created for SCCP endpoints, the max-dn command must
be configured in the telephom -sen ice mode. The maximum number of directory numbers
depends on the hardware platform ofthe Cisco Unified Communications Manager Express.
The max-dn command is set to zero b\ default to conserve system memory. If this value is set
above the required number of directory numbers, the router reserves system memory that it
could use for other functions.
Hie command max-dn ? helps to determine the maximum allowed number of ephone-dns that
the hardware supports. You should set the value within the range that complies with the license.
3-78 Implementing Cisco Voice Communications and QoS (CVOICE) vB.O 2010 Cisco Systems, Inc
Locale Parameters
This subtopic explains how toconfigure locale parameters inanSCCP environment.
Locale Parameters
user-locale [index] language-code
Specifies the language to be displayed on the phone
Multiple languages can be defined using the index (0-4)
Default: index 0. U.S. English
router (confIg-telephony) It
network-locale [index] language-code
Specifies the set of call progress tones and cadences on the phone
You can customize the Cisco Unified Communications Manager Express system with the local
language on the Cisco Unified IP phone display, as well as the call progress indicators and
cadence that the phone uses. This customization allows users to hear and interact with the
system using the language and audible cues that are familiar to them.
You can set the language that the phone displays and the call progress tones and cadences that
the phone uses to one of several ISO3166 codes that indicate specific languages and
geographic regions. The user-locale command specifics the language that the Cisco Unified IP
phone displays and the network-locale command specifies the set of call progress tones and
cadences that the phone uses.
The index command allows the configuration of up to five multiple user and network locale
settings. User/network-locale0 always holds the default setting that is used for all SCCP
phones that are not assigned alternative locales. The system default is US (United States),
unless a different locale is designated as default.
To apply alternative locales to different phones, you must use the cnf-Illes command to specify
per-phone configuration files. When you use per-phone configuration files, the configuration
file ofthe phone automatically uses the default locales in user locale 0 and network locale 0.
You can override this default for individual ephones by assigning locale tags to the alternative
language codes diat you want to use.
Date and Time Parameters
This subtopic describes howto define the date and time format for SCCPendpoints.
date - format {mm-dd-yy dd-mm-yy | yy-dd-mm , yy-mm-dd}
Sets the date format for phone displays
* Default, mm-dd-yy
router(config-telephony!#
time-format {12 24}
Selects a 12-hour or 24-hour clock for phone displays
Default. 12
k* 13 00 23.08.07
You can also modii> the fonnat in which the phone displays the date and time to the fonnat
that is typical for the location ofthe installation. You can use the date-format and time-format
commands to configure the date and time fonnat on a systemwide basis for all SCCP phones.
The following is a list of tvpical date formats that are supported by the date-format command:
dd-mm-y\: Sets the date to dd-mm-_\_\ fonnat
mm-dd-yy: Sets the dale lo mm-dd-\\ fonnat
yy-dd-mm: Sets the date to \\-dd-mm fonnat
yy-mm-dd: Sets the date to \\-mm-dd fonnat
fhe following is a list of Upieal lime formats that are supported:
12: Sets the time to 12-hour (a.m. and p.m.) format
24: Sets the time to 24-hour fonnal
implementing Cisco Voice Communications and QoS {CVOICE! v8.0 2010 Cisco Systems, Inc
Parameter Tuning
This subtopic explains howto tune additional parameters for SCCP endpoints.
Parameter Tuning
keepalive seconds
Sets the time interval between keepalives exchanged between the
Cisco Unified IP phones and Cisco Unified Communications
Manager Express.
The default is 30 seconds and the range is 10-65,535 seconds.
If three successive keepalives are missed, the device re-registers.
[codec {g711-ulaw | g722-64k}"
Selects the default codec for SCCP endpoints.
G.722-64K codec is supported on certain phones and firmware only.
If a phone does not support the codec, the default codec for that
phone is G.711 mu-law.
Default: G711-ulaw
Ihe keepalive time interval determines how frequently keepalives are sent between the Cisco
Unified IP phones and the Cisco Unified Communications Manager Express router. If the Cisco
Unified Communications Manager Express router fails to receive three successive keepalive
messages, it considers the phone to be out of service until the phone re-registers.
fhe default setting for the keepalives is 30 seconds. To change this interval, use the keepalive
command in telephony-service configuration mode.
Adjusting the keepalive determines how quickly that a failure is detected. To detect a failure
more quickly than 90 seconds, change the keepalive to a number lower than 30.
Note It may be useful to adjust the keepalive when phones register across a WAN link.
ITie codec command selects the default codec for SCCP IP phones in Cisco Unified
Communications Manager Express. The default codec is G.711 mu-law but it can be changed
to G.722-64k to help save network bandwidth for phones that are connected over the IP WAN.
fhe telephone finnware version on a Cisco Unified IP phone must support the specified codec.
If this command is configured and a phone does not support the specified codec, the default
codec for that phone is G.711 mu-law.
>2010 Cisco Systems, Inc. Cisco UnifiedCommunications Manager Express Endpoints Implementation 3-81
Generating Configuration Files for SCCP Endpoints
This subtopic explains how to generate configuration files for SCCP endpoints.
Generating Configuration Fifes for
ndDoii
create cnf-files
Builds the XMLconfiguration that is necessary for the phones
and stores them in the cnf-file location
router(config-telephony) #
Ienf - file {perphonetype perphone}
" Specifies the generation of different phone configuration files
by type of phone or by individual phone
router Iconfig-telephony)#
cnf-file location {flash: j slotO: / tftp tftp-url}
* Defines the storage location
Default: Asingle phone configuration stored in system
memory and used by all phones
Thecreate enf-filcs command in telephonv-sen ice configuration mode builds the XMI.
configuration files that are required for provisioning SCCPphones in Cisco Unified
Communications Manager Express. The command writes the files to the locationspecified with
the enf-fde location command.
The cnf-file {perphonetype | perphone} command in telephony-serviceconfiguration mode
affects how manv configuration tiles are generated using the create cnf-files command. Three
options exist as follows:
persystem: All phones use a single configuration file, fhis command is thedefault
behavior and therefore Cisco Unified Communications Manager Express does nol need this
command. The default user and network locale in a single configuration file arc applied to
all phones in ihe Cisco Unified Communications Manager Express system,
perphoneHpe: Creates separate configuration files foreachphone type, for example, all
Cisco Unified IP Phones 7965 use XMLDcfault7965.cnf.xml, and all Cisco Unified IP
Phones 7975 use XMLDefault7975.enf.xml. All phones ofthe same type use the same
configuration file, which isgenerated using thedefault userandnetwork locale. Thisoption
is not supported if the cnf-file location is configured for system.
perphone: Creates a separate configuration file for eachphoneby MAC address, for
example. SEP123456789.enf.xml. The configuration file for a phone is generated withthe
default user and network locale unless a different user and network locale is applied to the
phone using anephone template. Thisoption is not supported if the location option is
system.
The cnf-file location command in telephony-service configuration mode specifies a storage
location for phoneconfiguration files, fhe default is that a singlephoneconfiguration tile (pcr-
svstem) is stored in s\stem memory and is used by all phones. One of these locations can be
configured to store configuration files:
3-S2 Implementing Cisco VoiceCommunicationsand QoS (CVOICE) v8.0
System: This is the default. Whenthe systemis the storagelocation, there can be onlyone
default configuration file and it is used for all phones in the system. All phones, therefore,
use the same user locale and network locale.
Flash or slot 0: When flash or slot 0 memory on the router is the storage location, you can
create additional configurationfiles that can be applied per phone type or per individual
phone. Upto five user-defined user andnetworklocalescan be used in theseconfiguration
files. The generation of configuration files on flash or slot 0 can take up to a minute,
depending on the number of files that are being generated.
TFTP: When an external TFTP server is the storage location, you can create additional
configuration files that can be applied per phone type or per individual phone. Up to five
user-defined user and network locales can be used in these configuration files. TFTP does
not support file deletion. Whenconfiguration files are updated, they overwrite any existing
configuration files with the same name. If you change the configuration file location, files
are not deleted from the TFTP server.
>2010 Cisco Systems, Inc. Cisco Unified Communications Manager Express Endpoints Implementation 3-83
Cisco Unified Communications Manager Express SCCP
Environment Example
This sublopic provides an example of SCCP-related Cisco Unified Communications Manager
Express configuration.
Cisco Unified Commun
Express SCCP Environ
Loopback 0: 192.168.0.1
telephony-service
codec gT22-64)t
protocol mode dual-stactc preference ipv4
ip source-address 192.163.0.1 port 2000
user-locale 0 US
user-locale 1 BS
network-locale 0 US
network-locale 1 BS
time-format 24
date-format dd-nua-yy
keepalive 20
load 7965 SCCP45.S-0-2SR1S
cnf-file perphone
cnf-file location flash:
create cnf-files
max-ephoneB 200
max-dn 50 0
user-locale 0 US and network-locale 0 US do nol appear in Ihe configuration
Ihe figure shows a configuration example of systemwide SCCP parameters.
3-84 Implementing Cisco Voice Communicalions and QoS (CVOICE) vB0
Setting Up Cisco Unified Communications
Manager Express in a SIP Environment
This topic describes howto configure the Cisco Unified Communications Manager Express
system for SIP endpoints.
Configuring Global SIP Settings
router(config)#
voice service voip
Enters voice service voip configuration mode
router(conf-voi-serv)#
allow-connections sip to sip
Mandatory, allows calls between SIP endpoints
To configure the Cisco Unified Communications Manager Express to support SIP endpoints,
you need to enter the voice service VoIP configuration mode and allow calls between SIP
endpoints. using the allow-connections sip to sip command.
Configuring Global SIP
sip
Enters SIP mode
router(conf-aerv-sip)#
regis trar server
Mandatory, enables the SIP registrar server
- Default: SIP registrar server is disabled
router(conf-aerv-sip)#
bind {control media all} source-interface
Configures SIP interface binding feature
Must match source-address in voice register global mode
Only control option relevant
Further global SIP configuration is applied in the SIP mode. To enter the SIP configuration
mode, enter the sip command in the voice service VoIP mode, fhe SIP registrar server must be
enabled using the registrar sener command.
The bind command in SIP configuration mode binds the source address for SIP signaling or
media packets lo the IPv4 or IPv6 address of a specific interface, fhe binding of signaling
iraffie (control option) is relevant for the Cisco UnifiedCommunications Manager Express
support of SIP endpoints. because the media streams are terminated directly on Ihe SIP
endpoints. "fhis command is required if the Cisco Unified Communications Manager Express
should not use the IP lajer to dctennine the source IP address for SIP communications. If
configured, it must match the source-address command that is configured in the voiceregister
global mode.
Configuring Cisco Unified Communications Manager Express
for SIP
Thissubtopic describes how to configure themode of CiscoUnified Communications Manager
Express to support SIP endpoints.
Configuring Cisco Unified
Communications Manager Express for SIP
router(config)#
I voice register global
Enters voice register global configuration mode (system
settings of Cisco Unified Communications Manager Express
for SIP endpoints)
router(cooClg-raglster-globalI #
mode cme
Enters configuration mode for configuring SIP phones in
Mandatory for SIP-based Cisco Unified Communications
Manager Express
The voice register global command is used to set provisioningparameters fbr all supported
SIP phones in a Cisco Unified Communications Manager Express system.
The mode cme command enables Cisco Unified Communications Manager Express on the
router for configuration purposes. It should be issued before configuring SIP phones in Cisco
Unified Communications Manager Express to ensure that all required commands are available
in the configuration mode. The default setting is that the router is enabled only for Cisco SIP
Survivable Remote Site Telephony (SRST) but not for SIP-based Cisco Unified
) 2010 Cisco Systems. Inc Cisco Unified Communications Manager Express Endpoints Implementation
Configuring Source IP Address and Associating Firmware
This subtopic describes how to configure the source IP address ofthe Cisco Unified
Communications ManagerExpress for SIPcommunications and howto associate a phonetype
with SIP firmware.
'iguring Source IP
Associating Firmware
router(config-regieter-global)#
source-address ip-address [port port]
Optional, configures the source IP address and port
Must match the address defined by bind control source-
interface command in SIP mode
Default: TCP/5060, address provided by IP layer
router(config-regieter-global)#
load model firmware-file
Optional, associates a phone type with a phone firmware file
Adds the firmware specification into the configuration profile
Specifies firmware image without .loads suffix
1he source-address command in \oiee register global or cme configuration mode sets the
source address for communication with Cisco Unified Communications Manager Express SIP
endpoints. fhis command is required if the Cisco Unified Communications Manager Express
should not use the IP layer to determine the source IP address. If configured, it must match the
bind control source-interface command in SIP configuration mode.
The load command updates the configuration tile for the specified phone type to add the name
ofthe correct finnware file that the phone should load. This filename also provides the version
number for the phone finnware that is in the file. Later, whenever a phone is started up or
rebooted, the phone reads the configuration file lo determine the name ofthe finnware file that
it should load and then looks for that finnware file on the TFTP server. A separate load
command is needed for each type of phone.
for most Cisco Unified IP phones (including Cisco Unified IP Phones 7961. 7965. 7970. 7971.
and 7975) there are multiple finnware files. Eor these phones, use the TERMnn.x-y-x-w.loads
or SI Pnn,x-\-x-w, loads finnware filename for the load command, without the .loads file
extension. Eor such phones, you do not configure the load command for any finnware file other
than the TERM.loads or SIP.loads finnware file. In addition to the load command, use the (ftp-
server command lo enable TFTP access to the file by Cisco Unified IP phones. Ihe file
extensions are required when using the tftp-sener command.
Implementing Cisco Voice Communications and QoS (CVOICE) vS.O 2010 Cisco Systems, Inc.
Enabling SIP Endpoints
This subtopic describeshowto enable SIPendpoints in the Cisco Unified Communications
Manager Express environment.
Enabling SIP Endpoints
router(conflg-registar-global)W
max-pool max-phones
Mandatory, sets the maximum number of SIP phones
Default 0
router(config-register-global)#
Imax-dn max-directory-numbere
Sets the maximum number of SIP directory numbers
Default: 0
The mas-pool command that is configured in the voice register global configuration mode
limits the number of SIP phones (referred to as voice register pools) available in a Cisco
Unified Communications Manager Express system. The command is platform-specific. The
default value is 0.
'Ihe max-dn command that is configured in the voice register global configuration mode limits
the number of SIP phone directory numbers available in a Cisco Unified Communications
Manager Express system. The command is plattbnn-specific. The default value is 0. You can
increase the number of allowable extensions to the maximum, but after the maximum allowable
number is configured, you cannot reduce the limit without rebooting the router. You cannot
reduce the number of allowable extensions without removing the already configured directory
numbers with dn-tags that have a higher number than the maximum number to be configured.
12010 Cisco Systems. Inc. Cisco UnifiedCommunications Manager Express Endpoints Implementation
Locale Parameters
fhis subtopic describes bow to configure the locale parameters in the Cisco Unified
Communications Manager Express SIP environment.
Locale Parameters
user-locale [index] language-code
Optional, specifies the language to be displayed on the phone
Multiple languages can be defined using the index (0-4)
Default, index 0. U S. English
network-locale [index] language-code
Optional, specifies the set of call progress tones and cadences on
the phone
The locale parameters for a Cisco UnifiedCommunications Manager Express SIP environment
are configured identically to the SCCP environment, but in the voice register global
configuration mode.
The Cisco Unified Communications Manager Express system can be customized with the local
language on the display of SIP-based phones. The call progress indicators and cadence can also
be adjusted for SIP endpoints.
The user-locale command specifies the language that the Cisco Unified IP phone will display.
The network-locale command specifies the set of call progress tones and cadences that the
phone will use.
The index allows the configuration of up to i'ne multiple user and network locale settings.
User/network-locale 0 always holds the default setting that is used for ail SIP phones that are
not assigned alternative locales. The systemdefault is US (United States), unless a different
locale is designated as default.
3-90 Implementing Cisco Voice Communications and QoS (CVOICE) vS.O ) 2010 Cisco Systems. Inc.
This subtopic describes how toconfigure the date and time parameters in the Cisco Unified
Communications Manager Express SIP environment.
router(conflg-register-global)#
|date-format {m/D/Y | D/M/Y | Y/D/M | Y/M/D | YY/M/D}
Optional, sets the date format for phone displays
Default: M/D/Y
router(confIg-register-global)#
time-format {12 | 24}
Optional .selects a 12-hour or 24-hour clock for phone displays
Default: 12
timezone number
Optional, selects the time zone
Default 5 (Pacific Standard/Daylight Time)
The date and time parameters for the Cisco Unified Communications Manager Express SIP
environment are configured similarly to the SCCP environment, but they arc applied in the
voice register global configuration mode. The date-format and time-format commands are
used to configure the date and time format on a systemwide basis for all SIP phones.
The timezone command defines the time zone ofthe Cisco Unified Communications Manager
Express system.
NTP and DST Parameters
Ihis subtopic describes how to configure the NTP and daylight saving time (DS I) parameter;
in the Cisco Unified Communications Manager Express SIP environment.
NTP and DST Parametei
router(config-register-global) #
ntp-server ip-address [mode {anycast | directedbroadcast
multicas t unicast}]
Optional, configures the NTP server and the mode
router Iconfig-register-global)#
dst auto-adjust
* Enables daylight saving time (DST)
* Default enabled with default DST time period
dst {start stop} month [day day-of-month > week week-
number | day day-of-week] time hour:niinutes
Modifies the time period for DST
fhe ntp-sen er command specifics the IP address ofthe NTP server that is used by SIP phones
in a Cisco Unified Communications Manager Express system. It causes all SIP phones to be
synchronized lo the specified NTP sener.
The dst auto-adjust command enables the DSTadjustment of system time. It is enabled by
default with the default DST time period, which defines the summer time as the period from the
first Sunday of April at 2:00 a.m. (0200) until the last Sunday of October at 2:00 a.m. (0200).
The dst start/stop command is used to define the DSTperiod. It is only required, if it differs
from the default setting.
3-92 Implemenling Cisco Voice Communications and QoS (CVOICE) vS.O
Generating Configuration Files for SIP Endpoints
This subtopic explains how to generate configuration files for SIP endpoints.
Generating Configuration Files for SIP
Endpoints
.(or (eonfia-register-global) tt
create profile
Generates the configuration profile files required for SIP phones
Default: configuration files are notgenerated
router(conlq-reglBter-global)#
tftp-path {flaah: I alotOt | tftp://url>
Defines thedirectory towhich Cisco Unified Communications
Manager Express writes the configuration profiles for SIP phones
router(conflg-reqiBter-global)#
file text
- Generates ASCII text files oftheconfiguration profiles for SIP
phones
"Zl
The configuration files for SIP endpoints are referred to as configuration profiles, logenerate
the configuration profiles for SIP phones, use the create profile command in v01ce register
global configuration mode. The command generates the configuration tiles that are used tor
provisioning SIP phones and writes the files to the location specified with the tftp-path
command After achange to the SIP configuration files, it may be necessary to issue the no
create profile command to delete the existing file, followed by the create profile command to
re-create thefile, including thechanges just made.
fhe tftp-path command defines the directory to which the configuration profiles are written.
The default directory issystem memory (system:/cme/sipphonc/).
The file text command declares that the configuration profiles are written as ASCII text files.
Cisco Unified Communications Manager Express SIP
Environment Example
This subtopic prov ides aconfiguration example ofthe Cisco Unified Communications Manaee
f.xpress SIP em ironment.
Cisco Unified Communicati
Express SIP Environment E
Loopback 0: 192.168.0.1
allow-connectionu sip to sip
sip
bind control source-interface LoopbackO
oice register global
mode cms
source-address 132.16a.0.1 port 5060
user-locale 1 ES
uetworit-locaie 1 ES
time-format 24
data-format D/M/Y
timezone 13
ntp-server 9.9.9.9 mode directedbroadcast
load 7965 SIP45.9-0-2SR1S
tftp-path flash:
fila text
create profile
fhe figure provides aconfiguration example ofsystemwide Cisco Unified Communications
Manager Express SIPparameters.
ager
3-94 Implementing Cisco Voice Communicalions and QoS (CVOICE) vH.\
Summary
fhis topic summarizes the key points that were discussed in this lesson
Summary
. Cisco Unified Communications Manager Express supports
endpoints using Cisco proprietary SCCP, standard SIP, or
SIPwith Cisco Unified Communications extensions.
Endpoints usea setprotocols (PoE, Cisco Discovery
Protocol, DHCP, TFTP, SCCP, SIP) to communicate with the
CiscoUnified Communications Manager Express.
IEEE 802.3af standard differs from Cisco prestandard PoE by
offering higher power levels and optional parameters.
- Voice VLAN allows a separation of voice traffic from the data
traffic.
Summary (Cont.)
Cisco Unified IP phones receive this information via DHCP: IP
address/mask, default gateway, DNS, andTFTP (option 150).
NTP should beusedtosynchronize the time throughout the
entire Cisco Unified Communications infrastructure.
- Cisco Unified IP phones obtain via TFTP their specific ordefault
configuration, and the configuration-defined firmware.
SCCP-related parameters aredefined inthetelephony-service
configuration mode.
SIP-related parameters are configured in thevoice register
global {mode cme) configuration mode.
3-95
3-96 Implementing Cisco Voice Communications and QoS (CVOICE) 8 0 2010 Cisco Systems, Inc
Lesson 3
Configuring Cisco Unified
Express Endpoints
Overview
This lesson descnbes how to configure the Skinny Client Control Protocol (SCCP) and Session
initiation Protocol (SIP) endpoints in the Cisco Unified Communications Manager Express.
The SCCP endpoints are defined as the Ethernet phones (ephones) and have SCCP directory
numbers (ephone-dns) associated with them. The SIP endpoints are defined as voice register
pools and have SIP directory numbers (voice register directory numbers) associated with them.
This lesson discusses the various types of directory numbers available for Cisco Unified If
phones using either SCCP orSIP.
Objectives
Upon completing this lesson, you will be able to describe Cisco Unified Communications
Manager Express endpoint configuration elements. This ability includes being able to meet
these objectives:
Describe the types of directory numbers and how they are implemented in Cisco Uni fied
Communications Manager F.xpress
Explain how to configure directory numbers for SCCP phones
Describe how to define an IP phone type by configuring an ephone-type template
Explain how to configure major parameters of SCCP phones and how to assign directory
numbers to the phones
Explain how to configure directory numbers for SIP phones
Explain how to configure major parameters of SIP phones and how to assign directory
numbers to the SIP phones
Explain how to enable Cisco IP Communicator to register with Cisco Unified
Explain hou to generate configuration files for SCCP and SIP endpoints and how to reset
and restart SCCP and SIP phones
Describe hou to monitor and verify all major aspects of Cisco Unified Communications
Manager Express endpoint operation.
3-98 Implementing Cisco Voice Communications and QoS (CVOICE) vB.O 20I0 Cisco Systems Inc
Directory Numbers and Phones in Cisco Unified
r._i r. nnA Airoflnra nnnlhf
Phones and Directory Numbers
Name (SCCP)
Name (SIP)
What ills
Identifier
Number ot entities
aphone
voice register pool
Ptione as represented In
Cisco United Commuricatlons
Manager Expressconfigurafon
tag (sequence number)
Number of registeredendpoints
Phone can have one or more
directory numbersassociated
with it
Phone MAC address ties the
software configuration to tne
hardware
Impact onOial plan None
Directory Numtjer
ephone-dn
voce register *
Software coniguratlon thatrepresents the
lineconnecting a voice channel toapttone
dn-tag(sequencenumber)
Number ofsimultaneous calls{each irectory
number represent* avirtual voice port inthe
router)
Directory number can have one or more
telephone numbers associated with it
Drectory numbers areassigned totine
Buttons on phone*
Foreach directory number, one virtialvoice
port and one ormore dial peers arecreated
The phone represents the configuration and settings ofthe phys.eal IP phone and ,s associated
with aphysical device by MAC address. Aphone is configured as an SCCP ephone or
"Ethernet"phone." or avoice-register pool for SIP. The phone can be e.ther aCsco Unified IP
phone or an analog phone. Each phone has aunique tag, or sequence number, to identity ,t
duringconfiguration.
Adirector, number, also known as an ephone-dn for SCCP or avoice register dn for SIP is the
software configuration in Cisco Unified Communications Manager Express that represents the
line connecting avoice channel to aphone. Adirectory number has one or more extension or
telephone numbers associated with it to allow call connections to be made. Each directory
number has aunique dn-tag. or sequence number, to identify it during configuration. Directory
numbers are assigned to line buttons on phones during configuration. One virtual voice port and
one or more dial peers are automatically created for each directory number (one dial peer tor
each telephone number associated with the directory number) when the phone regtsters in Csco
Unified Communications Manager Express.
3-99
Directory Number Types
This subtopic describes the various types ofdirectory number
3-100
Directory Number Typ^
Single-line
directory number
Dual-line
directory number
Octo-line
directory number
Primary andsecondary extension B looiaraj
ona directory number
Shared directory number IB ioos j'
Multiple directory numbers withone "Wl 1QQ3
telephone number 5 11003
Overlaiddirectory
numbers
Ihe number of directory numbers affcels the number of simultaneous calls, because each
directory number represents a\ irtual voice port in the router.
The directory number is the basic building block of aCisco Unified Communications Manager
Express sv stem. Six dillerent types of directory numbers can be combined in different wavs for
different call coverage situations. The selection ofthe type depends on Ihc specific enterprise
requirements. For example, to keep the number of directory numbers low and prov ide serv ice
to alarge number ot people, shared directory numbers are useful. To have alimited quantity of
extension numbers and a large quantity ofsimultaneous calls, two or more directory numbe'rs
withthe sametelephone numbercan be created.
The directory numbers that are supported by the Cisco Unified Communications Manager
Express belong toany of these types:
Single-line (SCCP or SIP)
Dual-line(SCCP only)
Octo-line (SCCPonly)
Shared lineExclusive (SCCP onlv)
Shared-lineNonexclusive (SIPonlv)
Two directory numbers with one telephone number (SCCP orSIP)
Dual-number (SCCP or SIP)
Overlaid (SCCP only)
Implementing Cisco Voice Communications and QoS (CVOICE| v8 0
Single and Dual-Line Directory Numbers
Single and Dual-Line Directory Numbers
Charactensiics
Single-Une Directory DuaHfrie Directory;
iMmh Nurnaer
Voicechannels per button 1
Telephone numbers 1(SCCP), 1-10 (SIP) 1or 2primary and
^ secondary)
Lines dedicated to Lines supporting cal
liUrcom.poghg.MWI, waiting, caltransfer, or
and MOH conferencing
SCCP
Usage scenario
SIP/SCCP
Call to 1002
Endpoint support
Call lo 1001
^~p Answer
Concurrent Call to 1001
-^ Busy
Concurrent Call to 1002
Single-Line Directory Number: 1001
Dual-Line DirectoryNumber- 1002
Asingle-line directory number has the following characteristics:
. Supports one eall at atime using one phone line button. Asingle-line directory number in
SCCP has one telephone number associated with it. In SIP, it can have up to 10 telephone
numbers associated with it.
Ideal for lines that are dedicated to intercom, paging, Message Waiting Indicator (MWI),
loopback. and music on hold (MOH) feed sources.
Can be combined with dual-line directory numbers on the same phone.
Adual-line directory number has the following characteristics:
Onevoice port with twochannels.
Supported on Cisco Unified IP phones that are running SCCP, but not supported for SIP.
Can make two call connections at the same time using one phone line button. Adual-line
directory number has two channels for separate call connections.
Can have one number or two numbers (primary and secondary) associated with it.
Should be used for adirectory number that needs to use one line button for features such as
eall waiting. Call Transfer, orconferencing.
The figures demonstrate the difference between the single- and dual-line directory numbers
when asecond call is placed to them. The single line accommodates only one call and rejects
the second, while the dual-line can answer both, place one call on hold and take further actions,
such as Call Transfer and call conference.
Octo-Line Directory Number
This subtopic describes the features ofthe octo-line directory number.
3-102
Octo-Line Directory Number
Uptoeight activecallson a single bullon
!n|{?l^h?iPirZe>mmCOm'n9 Ca" fin9S a""""* <* a" *
!nT.?XZSôT65w1hout ac,ive calls-Phones lh
After acall is put on-hold, any phone that shares this directory number can
pick r! up
Supported on SCCPphonesonly
Call to 1001
o
Answer
Octo-L.ne Directory Number 10 1001
Call to 1001
Octo-Line DirectoryNumber 10. 1001
, **Z&^ Call-Wailing Tone ^W%& Answer
Oclo-Une Directory Number 10 1001 Octo-Line Directory Number 10 1001
Can resume either call put on Hold canresume either call put onhow
An octo-line directory number supports up to eight active calls, both incoming and outgoing on
asingle button. The octo-hne directory numbers are supported only on SCCP endpoints Unlike
adual-line directory number, which is shared exclusively among phones (after aeall is
answered, that phone owns both channels ofthe dual-line directory number) an octo-line
directory number can split its channels among other phones that share the directorv number All
phones are allowed to initiate or receive calls on the idle channels ofthe shared octo-line
directory number.
The figure demonstrates the operation ofan octo-line directory number. Because octo-line
directory numbers do not require adifferent ephone-dn for each active call, one octo-line
directorv number can process multiple calls. Multiple incoming calls lo an octo-line directorv
number ring simultaneously. The ringing slops when aphone answers acall. When phones '
share an octo-line directory number, incoming calls ring on phones without active calls and
these phones can answer. Phones with an active eall hear the call-waiting tone whenever a
subsequent call arrivesduringan aeliveconversation.
After aphone answers an ineoming call, the answering phone is in the connected state. Other
phones that share the octo-line directory number are in the remote-in-use state.
After aconnected call on an octo-line directorv number is put on-hold, any phone that shares
this directory number can pick up the held call. If aphone user is in the process of initialing a
Call Iransler or creating aconference, the eall is locked and other phones (hat share the octo-
line director) number cannot steal the call.
Implementing Cisco Voice Communications andQoS (CVOICE) v8.0
Nonexclusive Shared-Line Directory Number
fhis subtopic describes the nonexclusive shared directory numbers thai are used by SIP
endpoints.
Nonexclusive Shared-Line Directory Number
Multiple phones sharea common directory number
Up to 16callsat the sametime
Incoming call rings on all phones sharing the directory number
Calls put on hold notify the shared-line participants and can be
resumed by them
Supported by SIPendpoints
'. Call to 1001 xOfL . Anwar
Sri area-Line D.eclory Nimber 10: 1001 (SIP) ShareO-Une Directory Number 10: 1001 (SP)
'Call to 1001 j, jft^j Answer
^ %/ ^
SMtrt4.MDii.cio7Nimb.no 1001 (SIP) Shwd-Lln* Directory Number 10. 1001 (SIP)
Can rawme mucan put tnhold Car. rBSiime eitheraliput mhoH
Cisco Unified Communicalions Manager Express supports SIP shared lines to allow multiple
phones to share acommon directory number. All phones sharing the d.rectory number can
initiate and receive calls at the same time. Calls to the shared line ring simultaneously on all
phones without active calls. Any of these phones can answer the incoming calls The nnging
stops on all phones when aphone answers the call. The connected phones hear the call-waiting
tone onincoming calls tothe shared line number.
The phone that answers an incoming call is in the connected stale. Other phones that share the
directorv number are in the remote-in-use state. The first user that answers the call on the
shared Tine is connected to the caller and the remaining users see the call information and status
ofthe shared line.
Calls on ashared line can be put on hold like calls on anonshared line. When acall is placed
on hold other phones with the shared-line directory number receive ahold notihcat.on so that
all phones sharing the line are aware ofthe held call. Any shared-line phone user can resume
the held call If the call is placed on hold as part of aconference or Calllransfer operation, the
resume is not allowed. The ID ofthe held call is used by other shared-line members to resume
the call. Notifications are sent to all associated phones when aheld call is resumed on ashared
line.
Shared lines support up to 16 calls. The Cisco Unified Communications Manager Express
rejects any new call that exceeds the configured limit.
3-103
Exclusive Shared-Line Directory Number
This subtopic describes the exclusive shared directory numbers that arc used bv SCCP
endpoints.
Exclusive Shared-Line Directory Numb
Multiple phones share a common directory number
' Only one cal! at a time
Call appears on all phones sharing the directory number
Useful for answering a call at more than one phone
*Supported onSCCPendpoints
-^
call lo 1001
ShMd-lir-fl directory numbeMO ,01 jSccp) Snared-hnedirectoryrmrnberlQ 1001 (SCCP)
_ call to 1001
Sha'sa.vu directory numbirio 1001 (SCCP) Shared-line directofy number 10 1001 (SCCP)
Can resume either call put on hold Can resume either call put on hold
An exclusively shared directory number has the following characteristics:
Supported by SCCP endpoints only
I ine appears on two di ffercnt phones but uses the same directorv number, and extension or
phone number
Can make one call at atime: that call appears on both phones
Should he used uhen you want the capability to answer or pick up acall at more than one
phone
Because this directory number is shared exclusively among phones, ifthe direeiorv number is
connected to acall on one phone, that directory number is unavailable for calls on'any other
phone. If acall is placed on hold on one phone, il can be retrieved on the second phone This is
like having asingle-line phone in your house with multiple extensions. An incoming call can be
answered from any phone on which the number appears, and it can be picked up from hold on
anyphone on which ihenumber appears.
Implementing Cisco Voice Communications andQoS (CVOICE] v8 0
Multiple Directory Numbers with One Telephone Number
This subtopic explains bow multiple directory numbers can be configured with the same
telephone number.
Multiple Directory Numbers with One
Telephone Number
. Same telephone number on multiple separate virtual voice ports
Calls put on hold can be resumed only from tie same line
Useful for making more call connections with fewer numbers
. canbedual-line (SCCP only) or single-line directory numbers
<3>
1 Cal to 1001 <___, Answw
Directory NumbeMO 1001
2. Call to 1001 _y
Ditectoiy Number 10:1001
Can resume 1. cMIput on hoid
t^S
DirectoryNumber 11: 1001
iijj&k^ Answer
Directory Number 11: 1001
Caftrwume 2. cflllput on hold
Two directory numbers with one telephone number have the following characteristics:
Same telephone number that is combined with multiple separate virtual voice ports
supports multiple separate call connections
Can be dual-line (SCCP only) orsingle-line directory numbers
Can appear on the same phone on different buttons or on ditTerent phones
Suitable for making more calls while using fewer numbers
The multiple-directory numbers-with-onc-number situation is different from an exclusive
shared line (SCCP) which also has multiple buttons with one number but has only one
directorv number for all of them. An SCCP shared directory number will have the same call
connection at all the buttons on which the shared directory number appears. Ifacall on an
SCCP shared directorv number is answered on one phone and then placed on hold, the call can
be retrieved from another phone on which the shared directory number appears. But when there
are two directory numbers with one telephone number, acall connection appears only on the
phone and button at which the call is made or received.
3-105
Multiple-Number Directory Number
fhis subtopic describes the multiple-number directory numbers.
Directory number with multiple telephone numbers
Maximum two numbers (primary and secondary) in SCCP
Maximum 10 numbers in SIP
Useful when requiring multiple numbers for thesame button
with only one directory number:
One call with a single-line directory number (SCCP or
Two calls at the sametime using a dual-line directorv
number (SCCP)
Cal to 1001 or 1002 _ . ,_
Directory Number 101 1001.1002
Amultiple-number directory number has the following characteristics:
Maximum two telephone numbers (primary and secondary) for SCCP endpoints
Maximum 10 telephone numbers for SIP endpoints
Maximum one eall at atime ifit is asingle-line directory number
Maximum two calls at atime ifil is adual-line directory number (SCCP only)
Useful for multiple different numbers for the same button without using more than one
director* number
3-106 Implementing Cisco Voice Communications and QoS (CVOICE) i/8.0
2010 Cisco Systems. Ir
Overlaid Directory Number
This subtopic describes the overlaid directory numbers that are supported by SCCP endpoints
ffiiiiiiiilmill-^-B-iiir
Overlaid Directory Number
Up to25 directory numbers applied toone button
Up to 25 calls to the same phone number that resides on multiple phones
Acall placed on hold can be retrieved only by the phone that placed the
call on hold
All directory numbers must be either single-lineor dual-line (SCCP only)
<^
1. Call to 1001 Answer
One Button Directory Number 10 (1001),
Drectory Number!! (1002)
One Button- Directory Number 10 (1001),
Directory Number 11(1002|
^
2. Call to 1002
Can mswef He
(pulsfcaaontioW)
Answer
Ore Button. Directory Numbw 10 (1001), One Bunco; Directory Number 10 (1001).
Directory Number 11 (1002) Directory Number 11 (1002)
C- resume 1*ca8 putonhoM ortytMsplenecenresumsz^cafl
if it a War put on fM
An overlaid directory number has the following characteristics:
Is supported for SCCP endpoints only
Is amember ofan overlay set, which includes all the directory numbers that have been
assigned together toaparticular phone button
Can have the same telephone or extension number as other members ofthe overlay set, or
different numbers
Can be single-line or dual-line, but cannot be mixed single-line and dual-line in the same
overlay set
Can be shared on more than one phone
Up to25 lines can be overlaid on asingle button
Overlaid directorv numbers provide call coverage similar to shared directory numbers because
the same number'can appear on more than one phone. The advantage ofusing two directory
numbers in an overlay arrangement rather than as asimple shared line is that acall to the
number on one phone does not block the use ofthe same number on the other phone, which
would happen ifit were ashared directory number.
Cisco Unified Communicalions Manager Express Endpoints Implementation 3-107
Creating Directory Numbers for SCCP Phones
Ihis topic describes how to eoniigure directory numbers fbr SCCP endpoints.
Configuring Ephone-ds
router(config)#
jephone-dn dn-tag [dual-line | octo-line]
Creates an ephone-dn tor aphone line, an intercom line, a paging line
a vace-mail port, or at MWI
router(conElq-ephone-dn)#
number number [secondary number] [no-reg [both j primary]] |
Associates directory numberwithtelephonenumber.
The secondary numberoption defines a secondary number in addition to
the primary number
The no-reg keyword does not register these numbers with gatekeeper:
primary; Does not register pnmarynumber
both:Does not register primary nor secondary numbers
Without either option: Does not register secondary number
router (conflg-ephone-dn)#
name name
Sets name-based caller ID for calls originating from adirectory number
Ilie ephone-dn dn-tag global configuration command creates an ephone-dn, which builds one
\ irtual voice port. 1he dn-tag parameter must contain aunique number for anew ephone-dn or
an existing number ifyou are modifying acurrent ephone-dn. Ifyou want to assign (he ephone-
dn to an extension and aphone line, the ephone-dn needs to be able to accept two calls on the
same line at the same time. Use the keywords dual-line or octo-line at the end for the special
types ofdirectory number. Ifyou do not configure either option, the directory number is a
single-line directory number.
The number command defines avalid number for an ephone-dn that is to be assigned to an
SCCP phone. Ihe secondary key word allows \ ou to associate asecond telephone number with
an ephone-dn so that itcan be called by dialing either the main or secondary phone number.
The no-reg key word causes an H, 164 number in the dial peer to not register wilh the
gatekeeper. Ifyou do not specify both orprimary after the no-reg keyword, onlv the secondary
number does not register.
Anumber nomially contains only numeric characters that allow it (o be dialed from any
telephone keypad. However, in certain cases such as intercom numbers, which are normally
dialed only by the router, you can insert alphabetic characters into the number to prevent phone
users from dialing it and using theintercom function without authorization. Anumber can also
contain one or more periods (.) as wildcard characters that will match any dialed number in that
position. For example. 51.. rings when any four-digit number starling with 51 is dialed.
3-108 Implementing Cisco Voice Communications andQoS (CVOICE) v80
2010 Cisco Systems, inc
The name command is used to provide caller ID for calls originating from adirectory' number.
ms daiso generates local directory information that is accessed by ustng the
Directories button on aCisco Unified IP phone.
The name argument combination must match the order that is specified in the directory
L^mand (denned in te.ephony-se.ice mode): ^^TT^^r^\ôr
name string musl contain aspace between the first and second parts ofthe strtng ( first last
"last first").
>2010 Cisco Systems, Inc. Cisco Unified Communications Manager Express Endpoints Implementation 3-109
Single-Line Ephone-dn Configuration
This subtopic provides asample configuration for asingle-line SCCP directory number.
3-110
Single-Line Ephone-dn Configuration
One virtual voice port
* Only one call at anytime
6 Nosupport for call waiting
Useful for paging, intercoms, call parking slots MOH feeds
and MWI
ephone-dn 1
number 1001
One Virtual
Voice Port
One Channel
fhe figure illustrates aconfiguration example for asingle-line ephone-dn. The single-line
ephone-dn creates one \irtual port that supports only one channel at atime. It does not support
the call waiting feature and therefore Call Transfer and conferencing are not possible
Implementing Cisco Voice Communications andOoS(CVOICE) 8.0
20IOCisco Systems, Inc.
Dual-Line Ephone-dn Configuration
This subtopic provides asample configuration for adual-line SCCP directory number.
Dual-Line Ephone-dn Configuration
Twochannels, allowingseparate calls
Useful for calls that need call waiting, consultative transfer, and
conferencing on one button
Not recommended for intercoms, paging, MWI, call parking slots,
and MOH feeds (waste of resources)
ephone-dn 2 dual-line
number 1002
One Virtual
Voice Port
The figure illustrates aconfiguration example for adual-line ephone-dn. The dual-line ephone-
dn creates one virtual port that supports two channels. It supports the call waiting feature that
enables Call Transfer and conferencing. Dual-line ephone-dns are not recommended for
scenarios when the second channel is never used, such as for intercoms, paging, MWI, call
parkingslots, MOHsources.
Octo-Line Ephone-dn Configuration
This subtopic prov ides asample configuration for an octo-line SCCP directoiy number.
3-112
Octo-Line Ephone-dn Configuration
- One virtual voice port
Eight channels allowing separate calls
Useful for calls that need call waiting, consultative transfer,
and conferencing on one button
ephone-dn 3 octo-line
number 1003
Eight Channels
-<
One Virtual
Voice Port
1003 1003
I
1003 1003
f
1003 1003
1003 1003
Ihe figure illustrates aconfiguration example for an octo-line ephone-dn. The octo-line
ephone-dn creates one %irtual port that supports eight channels. The octo-line is useful when
eall coverage is implemented to ensure that calls are delivered to their intended destinations
Implementing Cisco Voice Communications andQoS(CVOICE) v8D
Dual-Number Ephone-dn Configuration
This subtopic provides asample configuration for adual-number SCCP directoiy number.
Duai-Number Ephone-dn Configuration
Primary andsecondary number
Support for one, two, or eight calls (single-line, dual-line,
octo-line)
number 1005 oecondary 2065559005
Two Channels
One Virtual
Voice Port
\
The figure illustrates aconfiguration example for adual-line ephone-dn with two telephone
numbers that arc configured on it. In SCCP, two is the maximum number of telephone numbers
that can be associated with one directory number. Calls to either number nng on this directory
number and can be answered on it. The number of concurrent calls depends on the type of the
ephone-dn.
3-113
Configuring SCCP Phone-Type Templates
^his topic describes the purpose and configuration of SCCP phone-type templates.
3-1M
SCCP Phone-Type Templates Introduction
Some endpoint types not supported in tnedefault Cisco Unified
Communications Manager Express settings.
CiscoUnified IP Phones 6901, 6911, wireless 7925
Expansion modules 7915,7916
Conference station 7937G
Third-party phones: Nokia E61
In such cases, phone-type templates are used to define the phone type.
Cisco Unified IPConference Station 7937G
Ciseo Unified Communications Manager Express classifies various endpoints bv their phone
type. Most ofthe phone types are predefined and can be referenced when configuring (he
deuces. The following phone types are not predefined within the Cisco Unified
Communications Manager F.xpress:
CiscoUnified IP Phones 6901. 69! I. and Wireless 7925
Cisco Unified IPPhone Expansion Modules 7915 and7916
Conference station: Cisco Unified IPConference Station 7937G
Third-part) phones: Nokia H6I
When an; ofthese endpoints exist in the network, the phone-type templates should be used to
define the phone t\ pe before it can be assigned tothe endpoints.
This subtopic explains how to configure SCCP phone-type templates
router(config)#
ephone-type phone-type faddon]
Create an SCCP ephone-type by defining anephone template
- phone-type: any alphanumeric string up to 32 characters
. addon option indicates that the type describes a module
router(config-ephone-type)#
[device-id number
Specifies thedevice ID forthe phone type
Device ID ispredefined for thespecific phone model
router<conflg-ephone-type)#
device-name name
Optional, assigns a nameto the phonetype
The ephone-type command creates an ephone-type template. Il defines aunique label that
identifies the type of phone. The label is any alphanumeric string up to 32 characters.
The addon option indicates that the phone type is an add-on module, such as the Cisco Unified
IP Phone7915Expansion Module.
The device-id command specifies the device ID ofthe type of phone being added with the
ephone-type template. If this command is set to the default value of 0, the ephone-lype is
invalid, fhe device IDs are preconfigured tothese values:
227: Cisco Unified IP Phone 7915 Expansion Module with 12 buttons
228: Cisco Unified IPPhone 7915 Expansion Module with 24buttons
229: Cisco Unified IP Phone 7916 Expansion Module with 12 buttons
230: Cisco Unified IPPhone 7916 Expansion Module with 24buttons
376: Nokia H61
484: Cisco Unified Wireless IP Phone 7925
431: Cisco Unified IP Conference Station 7937G
"Ihe device-name command isan optional command that allows the definition ofaname.
Configuring SCCP Phone-
(Cont)
router[config-ephone-type 1#
device-type phone-type
Specifies thedevice type for thephone
router(conflg-ephone-typeit
num-buttons number
Number of line buttons supported by the phone type
" numberRange: 1 to 100. Default: 0.
router (config-ephone-typel#
max-presentation number
Number of call presentalion lines supported by the phone type
numberRange: 1 to 100. Default: 0.
Ihe device type, num-buttons. and max-presentation commands are used to configure the
dev ice type, number of buttons, and number ofcall presentation lines that are supported bv the
phone type. These values are predetermined by the Cisco Unified Communications Manager
F.xpress andpresented in thefollowing table:
Supported Device
device-
id
device-
type
num-
buttons
max-
presentation
Cisco Unified IPPhone 7915 Expansion
Module with 12 buttons
227 7915 12
0 (default)
Cisco Unified IPPhone 7915Expansion
228 7915 24 0
Cisco Unified IPPhone 7916 Expansion
229 7916 12 0
CiscoUnified IPPhone 7916 Expansion
230 7916 24 0
Cisco Unified Wireless IP Phone 7925
484 7925 6 4
Cisco Unified IP Conference Station 7937G
431 7937 1 6
Nokia E61
376 E61 1 1
3-116 Implemenling Cisco Voice Communications andQoS (CVOICE) vS.O
Ephone Template for Conference Station 7937G Configuration
Example
This sublopic provides asample configuration for an ephone template.
Ephone Template for Conference
Station 7937G Example
Conference station type defined with an ephone-type template
ephone-type Conference7937
device-id 431
device-name Conference Station 7937G
device-type 7937
num-buttons 1
max-presentation 6
I
ephone 1
mac-address 001C.e21C.ED23
type Confarence7937
The figure shows asample template that defines the ephone type for the Cisco Unified IP
Conference Station 7937G. The type is then referenced by the SCCP device configuration.
3-117
Creating SCCP Phones
This topic describes how to provision SCCP endpoints within the Cisco Unified
Communications Manager hxpress.
Creating an SCCP
router(conflglB
ephone phone-tag
Creates anephone instance and enters ephone configuration mod
router(config-ephone)S
I mac-address mac-address
Associates the MAC address of the physical device with the ephom
IJ
The ephone command is used to create or modify an ephone. This command enters the ephone
configuration mode, where ephone-speeifie commands are issued.
1he mac-address command associates Ihe MAC address ofthe endpoint with the endpoint. It
specifies 12 hexadecimal characters in groups offour separated by periods. Ibr example
0000.0cl2.5456. P *
Configuring SCCP Ephone Type
This subtopic explains how to configure the ephone type.
Configuring SCCP Ephone Type
router (config-ephonal W . .
[type phone-type [addon 1 module-type [2 module-type]1
Optional, assigns a phone type toSCCP ephone
Identifies the appropriate firmware
Defines upto two expansion modules
Phone type can beeither predefined ordefined using
ephone-type template
12SP7902, 7905,7910. 7911, 7912.7920,7921,
7925 7931, 7935. 7936. 7937, 7940.7941,7941GE,
7942 7945, 7960. 7961, 7961GE, 7962, 7965, 7870,
7971! 7975. 7985, anl (analog), ata (CiscoATA-18B
orATA-188), bri(SCCP gateway),vge-phone
(VG248 analog phoneemulation)
7914,7915-12,7915-24,
7916-12,7916-24
n
The tvpe command in ephone or ephone-template configuration mode is used to assign aphone
type to an SCCP phone. It is not mandatory for ephone operations but it affects the
configuration file that is created for the defined endpoints and the default configuration file that
is generic to all phone types. In combination with the load command, il defines the firmware
image that should beused by aspecific phone model.
The addon option informs the router that an expansion module is added to the phone and
defines the type ofthe module.
The phone tvpes are preconfigured within the Cisco Unified Communications Manager Express
to the values shown in the table. Additional phone types can be defined using the ephone-type
templates. If the type command is applied both to the ephone-type template and to the ephone.
the value that isselinephone configuration mode has priority.
Configuring SCCP Ephone Buttons
This subtopic explains how to apply directory numbers lo SCCP phone buttons.
Configuring SCCP Ephont
routericonfig-ephone)#
button button-number {separator} dn-tag [,dn-tag...]
[[button-number {separator} dn-tag] [,dn-tag.. .]]
Associates ephone-dn with specific ephone button
Separator Description
Normal ring
Beep out no ring
Feature ring
Monitor mode for a shared line
Overlay set withoutcallwaiting
Overlay set withcallwaiting
Silent ring
Walch mode foral lines on a phone
Overlay rollover button
The button command allows aline button lo have one ormore ephone-dns assigned to it. The
button-number parameter represents the ph\ sieal phone button, with the top button being "1."
The separator parameter is asingle character that defines Ihe properties ofthe button:
: (colon): Normal ring. For incoming calls, the phone produces audible ringing, a (lashing
icon onihe phone display, and a Hashing red light onthehandset.
h: Beep but no ring, fhe audible ring is suppressed for incoming calls, but call-waiting
beeps are allowed. The visual cues are the same as those described for anormal ring.
f: Feature ring. This option differentiates incoming calls on aspecial line from incoming
calls on other lines. The feature ring cadence isatriple pulse, as opposed lo asingle pulse
for normal internal calls and a doublepulse for normal external calls.
m: Monitor modefor a shared line. Avisibleindicatorshows if theshared line is in use.
o: Overlay line without call waiting. Multiple ephone-dns share asingle bill ton. up toa
maximum of 10 on a button. The dn-tag argument can contain up to 10 individual dn-tag
values, separated bv commas.
c: Overlay line with call waiting. Multiple ephone-dns share asingle button, up toa
maximum of 10 on abutton. The dn-tag argument can contain up to 10 individual dn-tag
values, separated by commas.
s: Silent ring. The audible ring and the call waiting beep arc suppressed for incoming calls.
The visual cues arethe same asthose described for a norma! ring.
w: \\ ateh mode for all lines onthe phone for which this directory number isthe primary
line. Visible linestatus indicates whether the watched phone is idleor nol.
\: Creates ano\ erla\ rollover button. When the overlay button specified inibis command
is occupied b\ an active call, asecond call to one ofits ephone-dns will appear onthis
button, fhis button is alsoknown as anoverlay expansion button.
3-120 Implementing CiscoVoice Communications and QoS (CVOICEI v80
Configuring Ephone Preferred Codec
This subtopic explains how to configure the preferred codec for an SCCP endpoint
Configuring Ephone Preferred Codec
router(conflg-ephone)W
codec {g711ulaw [ g722r64 | g729r8 [diptarm-assist]
ilbc} .
Selects the preferred codec for the SCCP ephone
Affects calls to phones that are on thesame Cisco Unified
Communications Manager Express:
- Forother calls, the codec is negotiated usingVoIP
signaling protocols.
- For other calls, dspfarm-assist option enables DSP-farm
resources to be used for transcoding:
Ifresources unavailable, fall backto G.711 mu-law
Default: G.711 mu-law
The codec command is used to change the default G.711 mu-law codec to aless bandwidth-
intensive codec. G.722 (64 kb/s), G.729 (8 kb/s), or Internet Low Bitrate Codec (iLBC). The
firmware version ofthe telephone must support the specified codec. Ifacodec is specified by
using this command and aphone does not support the preferred codec, ihe phone will use the
global codec as specified by using the codec command in telephony-service configuration
mode. Ifthe global codec is not supported, the phone will use G.711 mu-law.
For calls to phones that are not in the same Cisco Unified Communications Manager Express
system (such as VoIP calls), the codec is negotiated based on the protocol that is used for the
eall (such as H.123). For calls to other phones in the same Cisco Unified Communications
Manager Express system, an IP phone that is configured to use G.729 will always have its calls
set up to use G.729.
When you use the codec command without the dspfarm-assist keyword, you affect only calls
between two phones on the Cisco Unified Communications Manager Express router (such as
between an IP phone and another IP phone orbetween an IP phone and aForeign Exchange
Station [FXS] analog phone). The command has no effect on acall that is directed through a
VoIP dial peer unless you use the dspfarm-assist keyword.
When vou use alow-bandwidth codec with the dspfarm-assist option, and the router isin a
VoIP call orconference that requires G.711, the digital signal processor (DSP) farm resources
are used toconvert G.711 tothe low-bandwidth codec. Adequate DSP resources must be
appropriately configured.
The benefit ofthe dspfarm-assist keyword is that itallows calls to use the G.729r8 codec,
which saves network bandwidth. The disadvantage isthe use of DSP resources that may be
required for other applications.
CiscoUnified Communications Manager ExpressEndpoinIs Implementation 3-121
Basic Ephone Configuration Example
This subtopic provides abasic ephone configuration example.
Directory number 7 provides one virtual port with two channels
Directory number 7 is applied to button 1
Phone needs tobe reset to use the settings
MAC0DOF.247O.F8F8
ephone-dn 7 dual lino
number 1001
ephone 1
mac-address 000F 2470 FBF8
type 7965
button 1:7
The example shows how to create ephone-dn 7and assign it to ephone 1. The ephone-dn
is configured as dual-line and is assigned to line button 1on the Cisco Unified IP phone with
the MAC address of000F.2470.F8F8.1 hephone type is 7965.
3-122 Implementing CiscoVoice Communications and QoS (CVOICE) vS.O
Multiple Ephone Configuration Example
This subtopic provides aconfiguration example with multiple ephones.
Multiple Ephone Configuration Example
epbone-do ID dual line
pumber 1004
ephone.-dn 11
number 1005
dual line
epbona-dn 12
number 1006
dual line
ephone 1
mac-address 000F 2470 FBF1
button 1:10
epbone 2
mac-address 0O0P 2470 A302
button 1:11
epbone 3
mac-address 0OOF 247 0 66F6
button 1:12
When there are multiple physical devices, the number of ephones must match the number of
devices. Then each ephone has one or more ephone-dns that are assigned to line buttons on the
physical device. This example shows three ephones with one ephone-dn that is assigned to each
of'them. Two phones use anormal ring while the third ephone uses the teature ring.
CiscoUnified Communications ManagerExpress Endpoints Implementation 3-123
Multiple Directory Numbers Configuration Example
This subtopic prov ides aconfiguration example with multiple directory numbers assigned lo
two ephones.
Example
1008 on burton 1
1009 on button 2
1010on bjlton 1
1011 on button 6
ephone -dn 14 dual -line
number 1008
ephone -dn 15 dual -line
number 1009
ephone dn 16 dual -line
number 1010
ephone dn 17 dual line
number 1011
ephone 5
mac-address 000F.2470.FAfil
button 1:14 . :15
ephone 6
mac-address Q00F.24 70.A7E2
button 1:16 6 :17
In the figure, multiple ephone-dns are assigned to the ephone buttons.
3-124 Implementing Cisco Voice Communications and OoS{CVOICE) vS.O
Shared Directory Number Configuration Example
This subtopic provides ashared directory number configuration example for SCCP endpoints.
Shared Directory Number Configuration
Example
1006 on button
1010 on button
1007 on button 1
1010 on button 6
One ephone-dn
applied lo two
different ephones
Only one phone
can use the
ephone-dn at a
time
Bolh phones ring
when a call
arrives, but only
one can pick up
the call
Eitherephone
can retrieve a call
placed on hold
ephone-dn 7 dual line
number 10OS
ephone-dn B dual line
number 1007
ephone-dn 9 dual line
number 1100
ephone 7
mac-address 000F 2470 FAA1
button 1:7 2:9
ephone 6
mac-address OOOF 2470 A7E2
button 1:6 6;9
The figure shows asample configuration for shared directory number in an environment with
SCCP endpoints. The exclusive (SCCP) shared ephone-dn has these characteristics:
It appears on two different phones, but uses the same ephone-dn and number.
Ifthe ephone-dn is connected to acall on one phone, that ephone-dn is unavailable for
other calls on the second phone because the phones share the same ephone-dn. The active
call appears on both phones.
You should use shared ephone-dns when you want die ability toanswer orpick up aeall at
more than one phone.
Both phones ring when acall arrives atthe ephone-dn but only one phone can pick up a
call, which ensures privacy.
When a call is placed on hold, eitherphone canretrieve it.
CiscoUnified Communications ManagerExpress EndpoinIs Implemenlation 3-125
Controlling Automatic Registration
This subtopic explains how to control automatic registration ofSCCP endpoints.
Controlling Automatic Rec
auto-reg-ephone
Enables automatic registration if phones not configured explicitly
Assoon as source IPaddress and maxephones are set
Default Enabled
router" sftow ephone
ephone-l[01 Mac;0024.C445.4B48 TCP sockets
REGISTERED in SCCP ver n/n max streanis=5
[1] aeti
mediaActive:0
reset:!) reset
IP:10.1.4.21
Preferred Cod
vhisper mediaActive:0
_sent:0 paging 0 debug
18443 796S keepaliv
ec: g711ulaw
startMedia
0 caps:9
0 max lin
:0 0
e 6
Lpcor Type: none
9:0 whisparLine:0
ofthook:0 ringing:0
The auto-reg-ephonc command allows automatic registration, inwhich Cisco Unified
Communications Manager Fxpress allocates an ephone slot toany ephone that connects toit,
regardless ofwhether the ephone appears in the configuration ornol. The auloregistration is
enabled by default.
The no fonn ofthis command blocks the automatic registration ofephones whose MAC
addresses are not explicitly listed in the configuration. When automatic registration is blocked.
Cisco Unified Communications Manager F.xpress records the MAC addresses ofphones that
attempt to register but cannot because they are blocked.
Use the show ephone attempted-registrations command toview the list ofphones that have
attempted toregister but have been blocked. Use the clear telephony-service cphone-
attempteci-registrations command toclear the list ofphones that have attempted toregister.
Implemenling Cisco VoiceCommunicalions and QoS (CVOICE) v8 0
Partially Automated Endpoint Deployment
This subtopic describes the partially automated SCCP endpoint deployment.
Partially Automated Endpoint Deployment
router(config-telephony)
auto aB8ign start-dn to stop-dn [type phone-type] [cfw
number timeout seconds] ^^^^^^^^^^^^^__^________m
Automated deployment of Cisco Unified IP phones
Automatically creates one ephone configuration for each available
directory number when a phone without aconfigured MAC address
registers
Before using this command, configure the ephone-dn tags tobe
assigned and defineat least one primary numberfor each dn-tag
- Requires that auto-reg-ephone command isenabled
All of the ephone-dns that you want todeploy must bethesame type
(single-line or dual-line).
Ephone-dns areautomatically assigned tonew ephones
The cfwandtimeout keywords define thecall forward number and
timeout values for phones that register
Multiple auto assign commands needed toassign discontinuous
ranges of ephone-dn tags and tosupport multiple phone types
The auto assign command in telephony-service configuration mode assigns ephone-dn tags to
SCCP phones as they register for service with the Cisco Unified Communications Manager
Express. This command enables you to assign ranges of ephone-dn tags according to the
physical phone type. You can use multiple auto assign commands to provide discontinuous
ranges and to support multiple types of IP phones. You can assign overlapping ephone-dn
ranges so that the ranges map to more than one type of phone. If there are not enough available
ephone-dns in the automatic assignment set, some phones will not receive ephone-dns.
If you do not specify atvpe in the auto assign command, the values in that range are assigned
to phones of anv type. If you do assign aphone type to aspecific range, the available ephone-
dns in that range are used first. The cfw and timeout keywords set the Call Forward Busy
(CFB) number and timeout values on all phones that automatically register.
The ephone-dn tags that the system automatically assigns must have at least aprimary' number
defined. All ofthe ephone-dns in asingle automatic assignment set must be ofthe same kind
(either single-line or dual-line). Automatic assignment cannot create shared lines.
Note The auto assign command grants telephony service to any endpoint that attempts to
register. The network should be secured against unauthorized access by unknown phones.
CiscoUnified Communicalions ManagerExpress EndpointsImplementation
3-127
Partially Automated Deployment Example
This subtopic provides an example ofthe partially automated SCCP endpoint deployment.
When a new IP phone registers, a new ephone is created with
the MAC address ofthe IPphone.
An existing ephone-dn is assigned tothe new epbone from the
ra nge defined for thetype of phone (starting with the lowest).
If a new IP phone does not match any auto assign command
with a type, theauto assigncommand without a type is used.
New Phone Plugs In
telephony-service
auto assign 1 to 10 type 7 961
auto assign 11 to 2D
type 1165
auto assign 21 to 40 type 7 975
auto assign tl to 50
number 1000
In the example, four auto assign commands declare different ranges ofephone-dn tags The
system will assign any Cisco Unified IP Phone 7961 the lowest unassigned ephone-dn from 1
to 10. The system will assign any Cisco IInified IP Phone 7965 (he lowest unassigned ephone-
dn from 11 to 20. and the system will assign any Cisco Unified IP Phone 7975 the lowest
unassigned ephone-dn from 21 to -40.
The directory numbers from the generic range of41 to 50 will be assigned to the specified
endpoints ifthey cannot be assigned an ephone-dn in the assigned range and to all unspecified
models of CiscoUnified IPphones.
3-128 Implementing Cisco Voice Communications andQoS (CVOICE) v8 0
Creating Directory Numbers for SIP Phones
This topic describes how to create directory" numbers for SIP endpoints
Configuring Directory Numbers for SIP
Endpoints
router(config)W
voice register dn dn-tag
Defines a directory number for a SIP endpoint (phone, intercom
line, voice port, or MWI)
router(config-regiater-dn)ft
I number number
* Defines a number for a SIP directory number
router(confIg-register-dn) #
|ahared-line [max-calla nmnber-of-callel
3
Creates a directory number to beshared by multiple SIPphones
max-calls option defines themaximum number ofcalls, both
incoming and outgoing
Range: 2 to 16. Default: 2
After the max-dn value is set toanondefault value, to enable the required number ofSIP
endpoints. the voice register dn command can be used to create directory numbers for SIP IP
phones directly connected in Cisco Unified Communications Manager Express. The command
defines adirectory number for aphone line, intercom line, voice-mail port, oran MWI. The
command also enters the voice register dn configuration mode, inwhich further parameters are
set.
The number command defines avalid number Tor an extension that istobe assigned toa SIP
phone. This command should be used before any other commands in voice register dn
configuration mode.
Anumber normally contains only numeric characters, which allows users to dial the number
from am telephone keypad. However, in certain cases, such as the numbers for intercom
extensions, the numbers can include alphabetic characters that can only bedialed internally
from theCisco Unified Communications Manager Express router and notfrom telephone
keypads. When alphabetic characters are included in the number, the extension can be dialed by
the router for intercom calls but notbyunauthorized individuals from other phones.
The shared-line command enables ashared line onanindividual SIP phone directory number.
The max-calls option defines the maximum number ofactive calls (from 2to16) allowed on
theshared line. If the shared-line command is not applied toa directory number, it does not
allow sharing by default. Ifthe shared-line command isconfigured without the max-calls
keyword, the directory number supports maximum two concurrent calls.
Voice Register Directory Number Configuration Example
This subtopic provides abasic configuration example for aSIP directory number.
3-130
lister
Useful for SIPphones, intercoms. MOH feeds, and MWI
voice register dn 1
number 1001
One Channel
One Virtual
Voice Port
I
II 1001 I
The figure illustrates how lo configure asingle-line directory number for SIP endpoints. The
voice register dn command creates one virtual port thai supports asingle voice channel. 'Ihis
configuration is useful forSIPphones, intercoms. MOH feeds, and MWI lines.
Implementing CiscoVoice Communications and QoS (CVOICE) v80
Creating SIP Phones
This topic describes how to define SIP endpoints within the Cisco Unified Communications
Manager Express.
Creating SIP Phones
router Iconfig)#
voice register pool pool-tag
Mandatory, defines a voice register pool that represents a SIP phone
router(con fig-registor-poo1)#
id {network address mask mask
address} ^^
ip address mask maak | mac
Mandatory, identifies a Cisco SIP IP phone
network and mask optionsenable it to accept SIP Register
messagesfrom anyphone within thespecified IP subnet
ipandmacoptions identify individual Cisco SIP IP phone
After the max-pool value is set to anondefault value, to enable the required number of SIP
endpoints in the Cisco Unified Communications Manager, the voice register pool command
can be used to create the SIP endpoints. The command enters the voice register pool
configuration mode, inwhich further parameters areset.
The id command explicitly identifies alocally available individual Cisco SIP IP phone in the
voice register pool configuration mode. This command must be used before any other
commands in the voice register pool configuration mode. This command offers adegree of
authentication, which isrequired to accept registrations, based on the following criteria:
Verification ofthe local Layer 2MAC address using the router Address Resolution
Protocol (ARP) cache. When the mac address keyword and argument arc used, the phone
must beinthe same subnet asthat oftherouter LAN interface, sothat the MAC address of
thephone is visible in therouter ARPcache.
Verification ofthe known single static IP address (or DI ICP dynamic IP address within a
specific subnet) ofthe SIPphone.
Configuring SIP Phones
This subtopic describes the configuration ofSIP phones.
Configuring SIP Phones
router(conflg-register-pool!#
type phone- type
Definesa phone type for a SIP phone
Supported types. 3951 (3911 and 3951). 7905,7906 7911 7912
7940.7941.7941GE,7942.7945.7960,7961 7961GE 7962
7965. 7970, 7971. 7975, ATA (ATA-186 orATA-186% P100 (PingTel
Xpressa 100). P600{Polycom SoundPoint 600)
router{config-regiBter-pool)#
number tag {number-pattern [preference value] [huntstop]
dn dn-tag}
* Mandatory, associates theSIP phone with a directory number.
Tag identifies thetelephonenumberif multiple exist Range is 1to 10.
Preferenceaffectstheorder There is no default.
huntstop specifes tostophunting if the number isbusy
The type command in voice register pool configuration mode defines aphone lype for a SI P
phone. The setting is required for the Cisco Unified Communications Manager Impress to write
the correct firmware specification into the configuration profile. The appropriate firmware is
found based on the phone type and the load command, which is set in the voice register global
configuration mode.
fhe number command in \oice register pool configuration mode indicates the L. 164 phone
numbers that are permitted b\ the registrar in the register message from the SIP phone. The
keywords and arguments ofthis command allow for more explicit setting ofuser preferences
regarding what number patterns should match the voice register pool. The tag identifies the
telephone number when there are multiple number commands (I to10 numbers are allowed).
The optional preference defines the preference order. Range is 0to 10, while the highest
preference is 0. The huntstop option stops hunting ifthe dial peer is busy.
3-132 Implementing Cisco Voice Communications and OoS(CVOICE) v80
Tuning SIP Phones
This subtopic explains how to tune the SIP phones.
Tuning SIP Phones
routerjconfig-regieter-pool)fl
I username username password string
Optional, defines authentication credentials for the SIP phone
Required if authentication isenabled in voice register global mode
router!config-register-pool)#
| dtmf-relay {[cisco-rtp] [rtp-nte] [sip-notify]}
. Optional, specifies a list ofnegotiated DTMF relay methods
. Default: None. DTMF tones remain inthe audio stream
router!conflg-regleter-pool)ft
codec a711alaw|q711ulaw|g722-64K|g729r8|ilbc
Optional, specifies thepreferred codecofthe SIPphone
Default: G.729 8 kb/s(differentfrom SCCPdefault: G.711 mu-law)
If both left by default SCCP-SIP call onsameCisco Unified
Communications Manager Express uses G.729
n
The username command invoice register global configuration mode sets the authentication
credentials for the SIP phone. It is used when authentication is required by the Cisco Unified
Communications Manager Express that is configured in the voice register global configuration
mode.
Tlie optional dtmf-relay command in voice register global configuration mode defines the dual
tone multifrequency (DTMF) relay methods that are supported by the SIP endpoint. This list ol
methods is advertised by the endpoint when negotiating the DTMF relay. By default, the
DTMF tones are transported within the Real-Time Transport Protocol (RTP) stream.
The optional codec command defines the preferred codec that is used by the SIP endpoint. The
default codec is G.729 8kb/s. The default codec ofSIP endpoints differs from the default codec
on SCCP endpoints (G.711 mu-law). An SCCP and aSIP endpoint that is registered to the
same Cisco Unified Communications Manager F.xpress communicate using the G.729 codec, if
bothendpoints use default codecvalues.
CiscoUnified Communicalions ManagerExpress Endpoints Implementation 3-133
Shared Directory Number Configuration Example
This subtopic prov ides anonexclusive shared directory number configuration example on SIP
endpoints.
Example
SIP shared mode
(non-exclusive)
Up to six phones
can use the
ephone-dn at a
time (max-calls)
Both phones ring
when a call
arrives one
answers the call
Either ephone
can retrieve a call
placed on hold
VOIC.
ihJr
r.glat
r 1100
sd-lins
ai dn J
calls 6
voico rojijt.r po
id mac 0D17.BO33
typ 7965
aumbec 1 dn 3
=1 10
voic. rogisti po
Id ipac 0OB1.CB13
typ 7965
tLiimbar 1 dn 2
1 11
0395
fhe figure illustrates the configuration ofadirecton number that isshared by two SIP
endpoints. fhe shared line supports amaximum ofsix concurrent calls, so even more endpoints
could beassigned tothisdirecton number. Thenonexclusive nature ofthe shared lineindicates
that the endpoints can make or recei\e independent calls. Aller acall in placed on hold, itcan
beretrieved b\ any phone that participates inthe sharing.
3-134 Implemenling CiscoVoice Communications and QoS (CVOICE] v8.0
Configuring Cisco IP Communicator Support
This topic describes how to configure the Cisco IP Communicator for interoperability with the
Cisco Unified Communications Managerfixpress.
installing Cisco IPCommunicator
Download Cisco IP
Communicator software
from http://www.cisco.com
Install the Cisco IP
Communicator on PC
Cisco IP Communicator is aWindows PC-based softphone application that allows the use ofa
PC to make premium voice and video calls. Offering the latest in IP communications
technology, itis easy to acquire, deploy, and use. With aUSB headset or USB speakerphone
and Cisco'lP Communicator, the users can easily access their corporate phone number and
voice mail.
To deploy the Cisco IP Communicator, the installation software should be first downloaded
from http://www.cisco.com and installed as prompted by the installation wizard.
Configuring Cisco IP Communicator
This subtopic explains how toconfigure theCisco IPCommunicator.
Configuring Cisco IP Communicat
Cisco IP Communicator settings (Menu >Preferences):
Network. Use these TFTP servers
- Audo Optimizefor low bandwidth (Yes. G.729, No: G711)
sa-emilY
AkaysCTop
ArjdDTurn; wtz*a
Qui* Search...
Cscounfiafl CMUsr Cpture...
is* Mwt jada \onetatos |
OMwHtW
" Ua NHKrt M laWBrm Bw>*m
Doric* hsne:
; Hwlsst |s^,-
tboui coco P CommuTicaD..
<" Ut>thishvuHunt
TFTPSarwr*
<" IM81*#JUTFTM
TFTP !*... t:
rFIPSorvarZ!
ItrKeHflaPfio/IWOFi.
WOTIMI55SSe
: Hmati* |s**:
" D(erPJnw' '
* |Souiv
P omen forkwhaviakth
For interoperability with Cisco U:nified Communications Manager Fxpress. the Cisco IP
Communicator needs the setting ofthe TFTP sener address. To set the TFTP address, navigate
toMenu >Preferences, select the Use These TITP Servers option, and configure the primary
and. optionalK. seeondan 1 FTP sener address.
The preferred codecthat is usedby the Cisco IPCommunicator is G.711 mu-law. In an
environment uith scarce bandwidth, the preferred codec can be set to G.729 8kb/s by checking
the Optimize for Low Bandwidth checkbox in the Preferences> Audio tab.
Implemenling Cisco Voice Communications and QoS (CVOICE;) v8.0
Managing Cisco Unified Communications
This topic describes bow to manage the SIP and SCCP endpoints in Cisco Unified
Communications Manager Express,
Rebooting Phones
Hard reboot
Must use for:
- Phone firmware
changes
- User locale changes
Network locale
changes
URL parameter
changes
DHCP and TFTP are
invoked
More time-consuming
restart
Soft reboot
Suitable for:
- Phone button changes
- Phone line changes
- Speed dial number
changes
DHCP and TFTP are not
invoked
Less time-consuming
When one or more phones that are associated with aCisco Unified Communications Manager
Express router are reconfigured, they must be rebooted to apply the new settings. Two
commands reboot the phones:
reset: The reset command performs ahard reboot that issimilar to apower-off, power-on
sequence. It reboots the phone and updates the phone with information from the DHCP
sener and TFTP server. This command takes significantly longer toprocess than the
restart command when you are updating multiple phones, but you must use the reset
command after updating firmware, user locale, network locale, orURL parameters.
restart: The restart command performs asoft reboot by simply rebooting the phone
without contacting the DHCP and TFTP seners. You can use the restart command for
simplebutton, line, or speed-dial changes.
Cisco Unified Communicalions Manager Express EndpoinIs Implementation 3-137
Rebooting Commands
Ihis subtopic presents the options for rebooting the phones.
Rebooting Commands
router(config-register-global]*
router(config-telephony|#
reset {all [time-interval] '
sequence-all}
restart {all [time-interval]
cancel | mac-address
i mac-address}
Resets or restarts one, several, or all phones
router(conflg-register-pool)#
routerIconfig-ephone]#
reset
restart
Resets orrestartsa single SCCP or SIPphone
ihe phones can be reset or restarted globally (telephony-service or voice register global mode,
respectively) or individually (ephone orvoice register pool, respectively). The individual reboot
affects only asingle device. The global reboot can specify the MAC address ofthe phone to be
rebooted and allows asequential reboot ofphones over time. The time interval (in seconds)
defines thetime between consecutive phone resets. Theinterval defaults to 15 seconds.
3-138 Implementing CiscoVoice Communications and QoS (CVOICE| v60
Verifying Cisco Unified Communications
This topic describes how to verify the proper operation of SIP and SCCP endpoints with the
Cisco Unified Communications Manager Express.
Verification Overview
Cisco Discovery Protocol
t - Verify that thephone received correct VLAN ID
I - Verify that the phone received correct IP address
I and scope options
Verify thatthecorrect files are inflash memory
tftp I Debug the TFTP server
Verify the installation ot the phonefirmware
Verify the phone status
DHCP
{:
Registration | .
Verify the received directory numbers

Verify that the locale is correct
The troubleshooting of endpoints commonly follows the same logical path that endpoints take
to register. The general sequence is defined by these steps, although some steps may not be
relevant in a given environment:
Verify the VLAN ID: The endpoint uses Cisco Discovery Protocol to obtain the Voice
VLAN from the attached switch. Use the Settings button onthe Cisco Unified IP phone to
check the VLAN configuration.
Verify the IP addressing: DHCP typically provides the IP parameters. Use the Settings
button onthephone tocheck theIP-related settings.
Verify the files in flash memory: Check and verify that the correct firmware files are in
the flash memory ofthe Cisco Unified Communications Manager Express router. This
infonnation is relevant for TFTP operations.
Debug the 'TFTP server: Ensure that the Cisco Unified Communications Manager
Express router is correctly providing the firmware and XML files via TFTP.
Verify the firmware installation on the phones: Use the Settings button on the phone to
check the firmware that the phone uses. The debug ephone register command on the
Cisco Unified Communications Manager Express also displays which firmware is being
installed.
Verify the phone status: Check the phone status using the Settings button on the phone,
\ iew the directory numbers that are assigned tobuttons and verify ifthe locale information
iscorrect on the endpoint. Verify the successful phone registration on the Cisco Unified
Verifying Phone VLAN ID
This subtopic explains how to verify the VLAN ID on the Cisco Unified IP ph
Settings > Network Configuration >Operational VLAN ID
TheVLAN IDreceded viaCisco Discovery Protocol from theswitch canbeviewed on the
Cisco Unified IP phone b\ pressing the Settings button and navigating toNetwork
Configuration > Operational VLAN ID.
3-140 Implemenling CiscoVoice Communications and QoS (CVOICE) v80
Verifying Phone IP Parameters
This subtopic explains how to verify the IP settings on the Cisco Unified IP phone
Verifying Phone IP Parameters
- Settings >Network Configuration >IPv4 Configuration
The IP parameters that are received via DHCP from the DHCP server can be viewed on the
Cisco Unified IP phone by pressing the Settings button, navigating to Network Configuration
>IPv4 Configuration and examining the IP Address, Subnet Mask, and Default Router
settings.
Cisco Unified Communications Manager Express EndpoinIs Implemenlation 3-141
Verifying Phone TFTP Server
This subtopic explains how to verify the received TFTP address on the Cisco Unified IP phone
Settings >Network Configuration >IPv4 Configuration
TFTP Serverl
The 1'FTP sener address that isreceived via DHCP option 150 from the DHCP server can be
\ ieued on the Cisco Unified IP phone by pressing the Settings button, navigating to Network
Configuration >IPv4 Configuration and examining the TFTP Server 1setting.
Verifying Firmware Files
This subtopic shows how to verify the firmware files on the Cisco Unified Communications
Manager F.xpress. ^
Verifying Firmware Files
Displays files in flash memory
jtec* Bho" flsah
4594326 Feb 26
531472 Feb 26
2160038 Fab 26
343039 Feb 36
1883455 Feb 21
6 42 Feb 26
642 Feb 26
642 Feb 26
69 Feb 26
4594326 Feb 25
531472 Feb 25
2582685 Feb 25
343039 Feb 25
1B8543S Feb 25
6 42 Feb 25
642 Feb 25
6 42 Fab 25
19 47 Feb 26
2010 13:14:50 Beep/apps45.9-0-2ES2.sbn
2010 13:17:04 scep/cmi45.9-0-2ES2.sbn
2O10 13:52:46 accp/=45sccp.9-0-2ES2.sbn
2010 13:55:02 sccp/dsp45.9-0-2BS2.sbn
2010 14:01:12 seep/3ar45scep.9-0-2ES2.sbn
2010 14:14:30 BOOP/SCCP45.9-0-2SR13.loads
2010 14:14:44 sccp/tero45. default .loads
2010 14:15 :00 seep/teriaSS. default, loads
2010 20:07:24 ayncinfo.nl
2010 16:59:28 appB45.9-0-2ES2.Sbn
2010 16:59:56 onu45.9-0-2E32.sbn
2010 17:01:00 cvn45sip.9-0-2BS2.sbn
2010 17:01:22 dap45.9-0-2E32.sbo
3010 17:01:54 Jar4Salp.9-0-2ES2.sbn
2010 17:02:12 SIP4S.9-0-2SR1S.loads
2010 17:02:32 tenn45.default.loads
2010 17:02:54 texm65.defsult.loads
2010 20:07:24 SIPDeEauIt.cot
The show flash command displays the contents of flash memory. The flash memory must
contain the fimiware files that are necessary for the Cisco Unified IP phone models that arc
deployed. Many other files can be in flash as well, depending on other configurations.
fhe figure shows that an SCCP firmware image for Cisco Unified IP phones 7945 and 7965
resides in the SCCP folder onthe flash memory. ASIP firmware image for Cisco Unified IP
phones 7945 and 7965 resides in the main directory ofthe flash memory.
Verifying TFTP Operation
3-144
This subtopic shows how to verify the TFTP service on the Cisco Unified Communications
Manager Fxpress.
Verifying TFTP Operati
Ensure that the SEP file for the phone is found
Examine if the correct firmware has been downloaded
loi
SEP0024C4455233.cnf.
:ned flash:/SEP 0024C4455233.on
lished flaah:/SEP0024C4455233.cn
Jking for SIP45-9-0-2SR1S.loads
aed flash:S1P45.9-0-2SR1S.loads
lished flash:SIP45.9-0-2SRlS.loai
.king for j ar45eip. 9-0-2ES2 .sbn
ned flaah:jar45sip.9-0-2ES2.sbn
,ished flash: jar45sip. 9-0-2ES2
king [or Cnu45 . 9-0-2ES2 .sbn
ned flash:cou45.9-0-2ES2.sbn, fd 10, aize 531472
ished flash:cnu45.9-0-2ES2.sbn. time 00:01:40
king for apps45.9-0-2ES2.sbn
ned flash:apps45.9-0-2ES2.sbn, fd 10,
iehed flaeh:apps45.9-0-2ES2,sbn. time 00:13:40
king for dep45.9-0-2ES2.sbn
ned flash: dsp45.9-0-2ES2.sbn. fd 10. size 343039
tug tftp e
37 :4i.S49:
37:44.853:
37 :45.357:
37:59.659:
37:55. 658:
merits
TFTP :
TFTP :
TFTP :
TFTP:
TFTP :
TFTP ;
TFTP:
TFTP :
TFTP :
TFTP j
TFTP :
TFTP :
TFTP:
TPTP :
TFTP :
TFTP :
TFTP :
.xml, fd 10.
38:00.390:
6:3S;00.894:
43:35.630:
43:40.570:
45:21.345:
45:23.277:
45 :23. 277 :
6:59:16.003:
fd 10
i. fd 10,
3 bn. t i me 00:05:34
Ihe debug tftp events command enables \ou to view output regarding files that are provided
b\ the TFTP server. You can view files, including finnware files, which arc specific toCisco
Unified Coninuinications Manager F\press to seewhether the Cisco Unified Communications
Manager Express router is using out-of-date or unsupported files. You can also view the XMI.
files for configured IP phones, the XMI. files for new IPphones, and the locale files.
Implementing Cisco VoiceCommunications and QoS (CVOICE] v8.0
Verifying Phone Firmware
This subtopic explains how lo verily the firmware image that the Cisco Unified IP phone is
running.
Verifying Phone Firmware
Settings >Model lnformation> Load File
The currently loaded firmware image can be viewed on tlie Cisco Unified IP phone by pressing
the Settings button, selecting Model Information, and examining the Load File information.
Verifying SCCP Endpoint Registration
This sublopic explains how to verify the successful registration ofSCCP endpoints.
Ityii
routeri show ephone
ephOQe-l[0J Kac:0024.C445.5233 TCP socket: [II activeLine: 0 -hisperLine0
REGISTERED in SCCP ver 19/17 max straans =5
nediaActive:0 whisper mediaActive, 0 startWedia:0 ofihook=o tin3ing;0 rese
reset_sent:0 paging 0 debug:0 caps:9
IP:10.1.2.13 - 53150 7965 keepalive 211 max line 6 avajlable line 6
button 1 : cw: 1 ecu : [0 0)
do 1 number 2001 CHI idle ch2 idle
button 2j cw: 1 cc: (0 0)
dn 3 number 2011 CHI IDLE CH2 IDLE
Preferred Codec; gTllulaw
Lpcor Type: cone
ephone-2[lj Kac:0024.C445.4B7P TCP Bocketj[2] activeLine:0 whiaperLine:0
RBGISTERED in SCCP ver 19/17 max streams.5
mediaActive:0 vhisper mediaActivei0 BtartMedia:0 ofhook:0 ringing;0 rese
reset_Bent:0 paging 0 debug:0 cape:9
IP:10.1.2.12 50439 7365 keepalive 211 max line 6 available line 6
button 1: c:l ccw:(0 0)
do 2 number 2002 CHI IDLE CH2 IDLE
Preferred Codec: gTllulav
Lpcor Type: none
The show ephone command is used to verif; ifthe phones have registered with the Cisco
Unified Communications Manager Fxpress router. Astatus of "registered" indicates that the
phonehas successfully registered. Astatusof-deceased'' indicates that there has beena
problem with keepalh es anda status of "unregistered'" indicates that theconnection wasclosed
normalk. The command display s the IP addresses and directory numbers that are assigned to
the endpoints.
3-146 Implementing Cisco Voice Communicationsand OoS (CVOICE) v8 0
Verifying SIP Endpoint Registration
This subtopic describes how to verify the successful registration ofSIP endpoints,
Verifying SIP Endpoint Registrati
router! show voice register all
VOICE REGISTER POOL
Pool Tag 1
Config:
Mac address is 0024.C445. 5233
Type is 7965
Number liat 1 -. DN 1
active primary line is: 1010
contact IP address: 10.1. 2. IS poit 5060
Dialpeers created:
session target ipv4:10.1.2 -IE: 5060
Statistics:
Active registrations : 3
Total SIP phones ragiater id: 1
Total Registration Statis tics
Registration requests : i
Registration success : 3
The show voice register all command displays all SIP-related Cisco Unified Communications
Manager Express configurations and register information. This information includes the
registration status ofall endpoints (voice register pools). To display the status ofasingle
endpoint theshow voice register pool command can beused.
>2010 Cisco Systems, Inc Cisco Unified Communications Manager Express Endpoints Implementation 3-147
Verifying SCCP Registration Process
This subtopic explains how to verify the registration process ofSCCP endpoints.
Verifying SGCP Registratioi
touterl debug epfcon e register
Mar 2 15:16:57.582 New Skinny socket accepted [11 (2 active)
Mar 2 15 i 16 : 57.5 B2 Bin family 2, Bin port 49692, in addr 10.90 .0.11
Mar 2 15:16:57.S82
skinny add socket 1 10.90.0.11 49692
Har 2 15:16:57.766
tIPPHONE-6-REG ALARM: 20: Name.SEP00OF2470P8F8
Load.3. 2 1,2. 14) L
ast'Phone-Keypad
Mar 2 15:16:57.766 Skinny StationAlarnMessage on socket [1] 10.90.0.11
SEPQO0F2470FSFS
Mar 2 15:16:57.766
20: Name*SEPOOOP2470FSFB Load-3 ,2 (2 .14 ) Las t-Phone- Keypad
Mar 2 15 : 16: 57.766 ephone-(11 [I] StationRegisterMessage (1/2/2) from
10. 90.0 .11
Mar 2 15 :16: 57.766
ephone-(1) [1) Register Stationldentifier DeviceNams
SEPOOOP2470 PBP8
Mar 2 15:16:57.766
ephone-!li (!) Statlonldentif ier Instance 1 deviceType 7
Mar 2 15 j16: 57.766
ephone-1 [-1] :stationIpAddr 10 .90. 0.11
Mar 2 15 :16: 57.766
ephone-1[1]:phone SEP000F2470F9F8 re-associate OK on
socket [1]
Mar 2 15:16:57.766
ÎPPHONE-6-REGISTER: ephone- 1:SEP000F2470F8FB
IP:10.90.3.i: ha registered.
Mar 2 15:16:57.766 Phone 0 socket 1
Mar 2 15 :16: 57.766
Skinny Local IP address - 10,95.0.1 on port 3000
Mac 2 15:16:57.766 Skinny Phone IP address . 10.90.0.11 49692
Mar 2 15 : 16: 57.766 ephone-1[11:Data Format M/D/Y
The debug ephone register command is used todebug the SCCP registration process onthe
Cisco Unified Communications Manager Express. After you have entered the debug ephone
register command. \ou may reset the phone and look lorthe Skinny StationAI armMessage text
inthedebug output, which should appear during thephone rcregistration process. The Load =
parameter should appear in thedisplay a few lines aftertheSkinny StationAlannMessage
output, followed byan abbre\ iatedversionnamethai corresponds to the correct finnware
filename.
Verifying SIP Registration Process
This sublopic explains how to verify the registration process of SIP endpoints.
Verifying SIP Registration Process
router* debug voice register events
Feb 26 20:18:12.143: voiCE_REG_POOL: Register request for U01O) from
(10.1.2.11)
Feb 26 20-18:12.143; VOICE RES POOL: Contact matches pool 1 number list 1
Feb 26 20:18:12.143: VOICE _RE0_P00L: keyflOlO) contact (10. 1.1. 11) and to
contact table
Feb 26 20:18:12.143: VOICE_RSG_POOL: key(lOlO) exists in contact table
Feb 26 20:18:12.1*3: VOICEBES POOL: contact (10.1.2 .111 added to contact table
Feb 26 20:18:12.147: VOICE RBG_POOL pool-tag 11) , dn->tag(ll. submasMl)
Peb 26 20:18:12.147: VOICE_RBG..POOL: Creating param container for dial-peer
40002. VOICE REGPOOL pool->tag(1). dn->tag(l), submaskfl)
VOICEREG POOL pooltag(l), dntag(l)
Feb 26 20:18:12.151: VOICP._RBG.PO0L: Crsated dial-peer entry of type 0
Feb 26 20:18:12.151: VOICE REG POOL: Registration successful for 1010,
registration id is 5
Feb 26 20:18:12.151: VOICE REG POOL: Pooltl]: service-control (reset type: 2)
message sent to flip:101010.1.2.IB
Peb 36 20:18:12.151: VOICE REG POOL: Contact matches pool 1 number list 1
Tlie debug voice register events command is used to debug tlie SIP registration process on the
Cisco Unified Communications Manager Express. This figure presents only a part oftheoutput
that is generated by the command, It includes information about the endpoint IP address
(10.1.2.11). the pool tag (1), the dn tag (1), and the telephone number (1010).
Verifying Endpoint-Related Dial Peers
This subtopic shows how to \erify the existence ofthe dial peers created by the registered
director*' numbers.
ing bndpoi
router* show dial-peer voice summary
dial-peer hunt 0
AD pre PASS OUT
TAG TYPE MIN OPER PREFIX DEST-PATTERN PER THRU SESS-TARGET STAT PORT KEEP.
20001 pots up up 10015 0 50/0/1
20002 pota up up 1002$ 0 50/0/2
40001 voip up up 1010 0 Byst ipv4:10.1.2.18:5060
router* show dial-peer voice 20001
peer type = voice, system default peer = FALSE, information type = voice.
description =
tag = 20001, destination-pattern = ~looi$>.
voice reg type * 0, corresponding tag > 0,
allow watch = FALSE
answer-addresa = "', preference^.
CLID Restriction = None
CLIO Network Number = ''
CLID Second Number sent
CLID Override RDNIS - disabled,
rtp-ssrc mux = system
source carrier-id , "-, target carrier-id . -'.
The show dial-peer voice summary command displays a summary ofdial peers in the system,
fhe list includes the SCCP endpoint and the SIP endpoint dial peers. TheSCCP-related dial
peers have tags in the range starting with 20001 and are shown as plain old telephone service
(POTS) dial peers. TheSIP-related dial peers have lags inthe range starting with 40001 and are
marked as VoIP dial peers. Specific information about agiven dial peer can he displayed using
the show dial-peer \ oice command with the relevant dial-peer lag.
3-150 Implementing Cisco Voice Communicalions and OoS (CVOICE! v8 0
Summary
This topic summarizes the key points that were discussed in th
lis lesson.
Summary
Directory numbers arevoice channels connecting tophones
and canbe configured as single-line, dual-line, octo-line,
shared lines, overlaid sets, and lines with multiple phone
numbers.
Directory numbers in SCCP phones are called ephone-dns
and support up toeight concurrent calls (octo-line).
SCCP phone type templates allow thedefinition ofephone
types that are not preconfigured in the Cisco Unified
Key ephone parameters include MAC address, button
configuration, ephonetype, and codec.
Summary (Cont.)
* Directory numbers ofSIP phones aresingle-line orshared
and can correspondto multiple telephone numbers.
- SIP phones areconfigured as voice register pools and
include an ID (MAC or IP), number, codec, and DTMF-relay
settings.
- Cisco IP Communicatorallowsthe configuration ofthe TFTP
server address and tuning the codec setting.
CiscoUnified IPphones can be hard-rebooted using the
reset command or soft-rebooted using the restart command.
Upon successful phone registration, the associated directory
numbers create virtual POTS dial peers.
Module Summary
Module Summary
Cisco UnifiedCommunications Manager Express is a
feature-richCisco IOSSottware-based call control platform
for up to 365 users,
Cisco Unified IP phones arethe most common Cisco Unified
Communications Manager Express endpoints and cansignal
calls using SCCP or SIP.
In Cisco Unified Communications Manager Express
configuration, SCCP directory numbers are ephone-dns,
SCCP phones areephones, SIP directory numbers arevoice
register dns, andSIPphones are voice register pools.
Thismodule introduces theCiscoUnified Communications Manager Express system, and
describes its key features and components. It lists the various types ofendpoints based on their
models and on the signaling protocol (Skinny Client Control Protocol [SCCP1 and Session
Initiation Protocol [SIP]). It explains the system parameters required when setting up Cisco
Unified Communications Manager Express, aswell asendpoint-related settings. The endpoint-
related configuration isgenerally divided into phone configuration (called ephones in SCCP
and voice register pools in SIP) and directory number configuration (called ephone-dns in
SCCP and voice register-dns in SIP). Special focus isgiven tothe various types ofdirectory
numbers supported on SCCP and SIP phones.
3-154 Implementing Cisco Voice Communications and QcS(CVOICE) v80 2010CiscoSyslems,Inc
Module Self-Check
solutions are found in the Module Self-Check AnswerKey.
Ql) Which statement correctly describes Cisco Unified Communications Manager Express?
(Source: Introducing Cisco Unified Communications Manager Express)
A) It targets small- tomedium-sized enterprises.
B) It targetsSOHOenvironments.
C) It supports upto465 endpoints.
D) Itprovides call-processing and voice-mail capabilities.
Q2) What is the key difference between Cisco Unified Communications Manager Express
and other Cisco Unified Communications Manager platforms? (Source: Introducing
Cisco Unified Communications Manager Express)
A) It provides additional features.
B) It is collocated with a voicegateway.
C) It ofTers a management CLI.
D) Il includes a voice-mail system.
Q3) Which two ofthese should be considered when migrating Cisco Unified
Communications Manager Express from aCisco 3800 Scries toa3900 Series
Integrated Services Router? (Choose two.) (Source: Introducing Cisco Unified
Communications Manager Express)
A) G2 (3900 Series) routers use generally the same licensing model as Gl (3800
Series) routers.
B) The current licensing model for Cisco Unified Communications Manager
Express is identical on G2andGl platforms.
C) "fhe number ofGl-supported endpoints isidentical for Cisco Unified
Communications Manager Express and SRST.
D) "fhe number ofG2-supported endpoints isidentical for Cisco Unified
Communications Manager Express and SRST.
E) G2 platforms require an activation license torun Cisco Unified
Q4) Hou does Cisco Unified Communications Manager Express provide reachability ofits
registered endpoints to external callers? (Source: Introducing Cisco Unified
Communications Manager Express)
A) It intercepts signaling exchanges and forwards theappropriate call setup
requests lo its endpoints.
B) Itregisters the numbers on avoice gateway that isinthe voice path.
C) It distributes theendpoint numbers among neighboring gateways.
D) Itautomatically creates virtual dial peers that appear in the gateway dial plan.
05) Which of these phones feature softkeys? (Source: Examining Cisco Unified
Communications ManagerExpress Endpoint Requirements)
A) basic Cisco Unified IP phones
B) midrange Cisco Unified IP phones
C) upper-end Cisco Unified IP phones
D) basic, midrange. and upper-end Cisco Unified IPphones
E) midrange and upper-end CiscoUnifiedIPphones
i2010 Cisco Systems, Inc Cisco Unified Communications Manager Express Endpoints Implementation 3-155
Q6) Put the steps ofthe Cisco Unified IP phone boot process in the correct order. (Source:
Examining Cisco Unified Communications Manager Express Endpoint Requirements)
_ requesting and obtaining the configuration file
... registering on Cisco Unified Communications Manager
_. obtaining power from the switch
requesting L)l ICP infomialion
requesting and obtaining the firmware tiles
configuring VLAN
obtaining the DHCP infonnation
Q7) What are two differences between Cisco prestandard PoE and IEEE 802.3af? (Choose
two.) (Source: Examining Cisco Unified Communications Manager Express Endpoint
Requirements)
A) Cisco prestandard PoE uses East Link Pulses.
B) II EE 802.3af delivers power only todevices that require it.
C) Ciscodevices requireCiscoprestandard PoE.
D) Pins that are used in Cisco prestandard PoE (1,2, 3. 6) arc incompatible with
IEEE 802.3af.
E) Cisco prestandard PoE does not classify power levels.
Q8) Which two ofthese should you do todeploy the voice VLAN feature'.' (Choose two.)
(Source: Examining Cisco Unified Communications Manager Express Endpoint
Requirements)
A) Configure voice and auxiliary VLANs onthe ports with attached IP phones.
B) Enable the PortEast feature to shorten theinitial latency.
C) Configure 802, IQtrunking onthe switch ports with attached IP phones.
D) Configure 802. IQor ISL trunking onthe switch ports with attached IP phones.
E) Allow DHCP to ensure that the voice VLAN 11) is advertised.
E) Use Cisco switches.
Q9) DHCP option is used in telephony environments witha Cisco Unified
Coninuinications Manager platform to direct the booting phonesat the TFTPserver.
(Source: Examining Cisco Unified Communications Manager F.xpress Endpoint
Requirements)
QIO) Which two sen ices require clock accuracy? (Choose two.) (Source: Examining Cisco
Unified Communications Manager Express Endpoint Requirements)
A) RTP sequencing
B) RTCP delivery monitoring
C) eall lists
D) NTP
E) CDRs
3-156 Implementing Cisco Voice Communications andQoS (CVOICE) vS.O <)2010 Cisco Systems, Inc
Ql 1) When does aphone request afirmware image? (Source: Examining Cisco Unified
Communications Manager Express Endpoint Requirements)
A) ifit does not receive its specific configuration file SEP<m<ic>.cnf.xml
B) ifitreceives the generic configuration file XMLDcfault.cnf.xm! with asetting
that is different from the current image
C) ifitreceives its specific configuration file with the required image information
embedded in it
D) ifthe generic configuration file is missing and the specific file is received
Q12) Which ofthese must be configured to activate Cisco Unified Communications Manager
Express for SCCP endpoints? (Source: Examining Cisco Unified Communications
ManagerExpress Endpoint Requirements)
A) telephony-service command
B) nothing: Cisco Unified Communications Manager Express enabled for SCCP
phones by default
C) generation of configuration files
D) source IP address
E) protocol mode: IPv4, IPv6, or both
Q13) Which two ofthese must be configured toactivate Cisco Unified Communications
Manager Express for SIP endpoints? (Choose two.) (Source: Examining Cisco Unified
Communications Manager Express EndpointRequirements)
A) voice register global command
B) generation of configuration files
C) source IP address
D) operation mode: CME or SRST
E) SIP registrar server
Q14) What arc two types ofephone-dns that are available in aCisco Unified
Communications Manager Express system? (Choose two.) (Source: Configuring Cisco
Unified Communications Manager Express Endpoints)
A) single-line ephone-dn
B) secondary andtertiary extension ononeephone-dn
C) shared ephone
D) multiple ephoneson one ephone-dn
E) overlay ephone-dn
Q15) Which command creates an ephonc-dn that builds one virtual voice port? (Source:
Configuring Cisco Unified Communications Manager Express Endpoints)
A) router(config-telephony)# ephone-dn dn-tag
B) routcr(config-telephony)# number dn-number
C) router(config)# ephone-dn dn-tag
D) router(config)#ephone-dn dn-number
Q16) Which two phone types should be created using the SCCP phone template? (Choose
two.) (Source: Configuring Cisco Unified Communications Manager Fxpress
Endpoints)
A) Cisco Unified IP Conference Station 7937G
R) Cisco Unified Wireless IP Phone 7925
C) Cisco Unified IP Phone 7961
D) Cisco Unified IP Phone 7965
E.) Cisco Unified IP Phone 7975
QI7) Which two configuration commands (max-ephoncs. type, button, mac-address) are
mandatory to create an SCCP endpoint? (Choose two.) (Source: Configuring
Cisco Unified Communications Manager Express Endpoints)
Q18) Which two types ofdirectory numbers exist for SIP endpoints? (Choose two.) (Source:
Configuring Cisco Unified Communications Manager Fxpress Endpoints)
A) single-line
R) dual-line
C) oeto-Hne
D) shared directory number
E) o\er!aid directory number
Q19) How can you configure calls between SCCP and SIP endpoints that are registered to
thesame Cisco Unified Communications Manager Express louse theG.711 a-law
codec0 (Source: Configuring Cisco Unified Communications Manager Express
Endpoints)
A) codec g711ala in telephony -service configurationmode
B) codec g71Iahm in ephone configuration mode
C) codec g71lalaw in ephone-dn configuration mode
D) codec g711alaw in voiceregisterglobal configuration mode
E) codec g711alaw in \ oice registerpool configuration mode
I) codec g711alaw in voiceregisterdirectory number configuration mode
Q20) Which twoprotocols aresupported byCisco IPCommunicator? (Choose two.)
(Source: Configuring Cisco Unified Communications Manager Express Endpoints)
A) RTP
B) cRTP
C) DHCP
l CTIQBE
E) SIP
Q21) Which command performs a hardreboot, similarto a power-off, power-on sequence?
(Source: Configuring Cisco (inifiedCommunications ManagerExpress EndpoinIs)
A) restart
B) reset
C) reload
D) restart, reset, or reload
E) either restart or reset
3-158 Implementing Cisco VoiceCommunications and QoS (CVOICE) v8 0 2010 Cisco Systems, Inc.
Q22) Which procedure verifies that the DHCP server is providing the correct information to
the IP phones? (Source: Configuring Cisco Unified Communications Manager Express
Endpoints)
A) issuing the show running-config command
B) issuing the show flash command
C) issuing the debug ephone register command
D) pressing the Settings button, and then, from the menu that appears, choosing
Network Configuration settings
Q23) Which procedure is used to monitor the boot sequence ofan SCCP endpoint? (Source:
Configuring Cisco Unified Communications Manager Express Endpoints)
A) issuing theshowrunning-config command
B) issuing theshowflash command
C) issuing the debugephone register command
D) pressing the Settings button, and then, from the menu that appears, choosing
Network Configuration settings
2010Cisco Systems, Inc Cisco Unified Communicalions Manager Express EndpoinIs Implementation 3-159
01 I A
021 IJ
0?> 13. C
041 D
05) D
QO) I obtaining power fiom the switch
2 configuring VLAN'
j* requesting DHCP information
4, obtaining theDHCP infomialion (including option 150)
^ requesiing and obtaining the configuration file
6 requestmy and obtaining [lie finnware file>
7 registering on Cisco Unified Communications Manager
Q7) A. 1
QSt B. F-
(.19) 150
QlOi C. I:
Qll i li
yi:i D
Oi-'i L. I
014 i -V 1
01?)
C
Qlo) A. 15
01 "t mav-ephones. mac-address
018) a. n
019)
F
Q2l)i A.F
Q2H B
Q22) [J
02}) C
Module 4
Dial Plan Implementation
Overview
Dial plans are essential for any Cisco Unified Communications deployment. Whether you are
implementing single-site ormultisite deployments, having athorough understanding ofdial
plans and the knowledge ofhow to implement them on Cisco IOS gateways isessential for any
engineer who designs and implements aCisco Unified Communications network.
This module describes thecharacteristics of a dial plan, associated components on CiscoIOS
gateways, andexplains howto implement dial plans.
Module Objectives
Upon completing this module, you will be able todescribe the components ofadial plan and
explain how toimplement adial plan onaCisco Unified voice gateway. This ability includes
Describe the characteristics and requirements of a numbering plan
Explain thecomponents of a dial planandtheirfunctions
Describe howto configure a gateway for digit manipulation
Explain howto configure a gateway toperform pathselection
Describe how lo configure calling privileges on a voice gateway
4-2 Implementing Cisco Voice Communications and QoS (CVOICE) vS.O 201C Cisco Systems, Inc.
Lesson 1
Introducing Call Routing
Overview
Tointegrate VoIP networks into existing voice networks, youmust have theskills and
knowledge toimplement call routing and design anappropriate numbering plan. Ascalable
numbering plan establishes thebaseline for a comprehensive, scalable, and logical dial plan,
Thislesson describes call routing principles, discusses attributes of numbering plansforvoice
networks, addresses thechallenges ofdesigning these plans, and identifies themethods of
implementing numbering plans.
Objectives
Upon completing this lesson, you will beable todescribe how a gateway routes calls and how
to implement a numbering plan. This ability includes being able to meet these objectives:
Describe the basiccharacteristics of a typical numbering plan and list the different types of
numbering plans
Explainthe attributes of a scalable numbering plan
Describe overlapping numbering plansandstrategies to address the issue of overlap
Describe howto integrate privateand public PSTN numbering plans
Explain howa gateway implements the numbering plan
Describe how a voice gateway routes calls
Introducing Numbering Plans
4-4
This topic introduces numbering plans and lists the most common public numbering plans.
lumbering Plans
1Anumbering plan is a numbering schemewith thefollowing
characteristics:
Defines a set of rules to allocate numbers used in
telecommunications
Is based on telecommunications standards
Is established by numbering plan authorities, which
regulate the distribution of numbers and codes in their
territory
Many regional and national numbering plans exist'
NorthAmerican Numbering Plan (NANP)
European Telephony Numbering Space (ETNS)
Many countries have their own numbering plans
A numbering plan is a numbering scheme that is used in telecommunications to allocate
telephone number ranges to countries, regions, areas, andexchanges, andlo nonfixed telephone
networks such as mobile phone networks. Thenumbering plandefines Ihe rules for assigning
numbers to a device.
Tjpes of numbering plans include the following:
Private numbering plans: Private numbering plans are used to address endpoints and
applications within privatenetworks. Private numbering plans are not required to adhereto
any specific format and can be created to accommodate the needs ofthe network. Because
most private telephone networks connect to the public switched telephone network (PSTN)
al some point in the design, it is good practice to plan the private numbering plan to
coincide with publicly assignednumber ranges. Number translation may be required when
connecting private voice networks lo the PSTN.
Public or PSTN numbering plans: PSTN or public numbering plans are unique to the
country in which they are implemented. The most common PSTN numbering plans are
explained in this topic.
Implemenling Cisco Voice Communications and QoS (CVOICE) v8 0 2010 Cisco Systems, Inc.
few
North American Numbering Plan
This subtopic describes the North American Numbering Plan (NANP).
North American Numbering Plai
NANP is the numbering plan for the United States and its
territories, Canada, Bermuda, and many Caribbean
countries.
It is administered by the NANPA.
NANPnumbers are 10-digitnumbers.
Local Number;
<2-9>XX- CO Code
XXXX = Line Number Area Code
/
^2-9>XX*=2-9>XX-XXXX
X = <09>
The NANP is an integrated telephone numbering plan that serves 19North American countries
that share its resources. Developed in 1947 andfirst implemented in 1951 byAT&T, theNANP
simplifies and facilitates ihe direct dialingof long-distance calls. The countries that use the
NANP include the United States andits territories, Canada, Bermuda. Anguilla, Antigua and
Barbuda, the Bahamas, Barbados, the BritishVirginIslands, the CaymanIslands, Dominica,
the Dominican Republic, Grenada, Jamaica, Montserrat. St. Kitts and Nevis, St. Lucia. St.
Vincent and the Grenadines. Trinidad and Tobago, andTurks and Caicos Islands.
NANPnumbers are 10-digit numbers, usually formatted as follows, NXX-NXX-XXXX. in
whichNis anydigit from 2 to 9, inclusive, and Xis any digit from0 to 9, inclusive.
'[Tie first three digitsof anNANP number (NXX) arecalled theNumbering Plan Area(NPA)
code, oftencalledthe area code. Thesecondthreedigits (NXX)arc calledthe centraloffice
(CO) code, switched code, or prefix. Thefinal four digits (XXXX) arecalled the linenumber or
station number.
The NorthAmerican Numbering PlanAdministration (NANPA) administers the NANP.
NANP Numbering Assignments
This subtopic describes the NANP numbering assignments.
NANP Numbering Assignments
The NANP includes the following conventions:
Easily Recognizable Codes (ERCs)
Second and third digits are the same, such as 800 and
877 toll-free numbers
- Carner Identification Codes (CICs):
4-digit codes used to route and bill calls
International dialing1
011 followed by the country code and country-specific number
Long distance
1-<2-9>XX-<2-9>XX-XXXX
In-statelong-distance or local call
<2-9>XX-<2-9>XX-XXXX
Local call
<2-9>XX-XXXX
Thearea that is served by theNANP is divided intosmaller areas, each identified bya three-
digit NPA code, orarea code, fhere are 800 possible combinations ofarea codes with the NXX
format. Ilowe\ er. some of these combinations are not available or have been reserved for
special purposes, asshown inthe NANP Numbering Codes table.
NANP Numbering Codes
Reserved Code Purpose
Easily Recognizable
Codes (ERCs)
When the second and thirddigitsof an area code are the same, that code is
called an ERC These codes designate special use, such as toll-freeservice
Automatic Number
Identification (ANI] II
digits
ANI II digitsare two-digit pairs that are sent with the originating telephone
number as part ofthe signaling that takes placeduring the setup phase of a
call These digits identifythe type of originatingstation.
Carrier Identification
Codes(CICs)
CICs are used to route and bill calls in the PSTN. CICs are four-digit codes in
the format XXXX, where X is any digitfrom 0 to 9, inclusive. There are separate
CIC pools fordifferent featuregroups, such as line-side and trunk-side access.
International dialing
Youdial 011 beforethe countrycode and the specificdestinationnumber to
signal that youare placingan international call.
Long distance
The first "1"dialed defines a toll call within the NANP, where Xis a digit from0
through 9
In-state long
distance or local call
A10-digit number may either be a toll call within a common region or, inmany
larger markets, may bea local call ifthe area codeisthe same as the source.
7-digit number (<2-
9>XX-XXXX)
A7-digit number defines a local call. Some larger areasuse 10-digit numbers
instead of 7-digit numbers to define localcalls
Implemenling Cisco Voice Communications andQoS(CVOICE) v8.0
NANP Numbering Assignments (Cont)
N11 codes:
- Service codes: 911 emergency. 511 traffic, 411 information, and
so on,
456-<2-9>XX-XXXX codes
- Carrier ID for inbound international calls
- <2-9>XX defines the earner
555-01XX line numbers:
- Reserved for fictitious uses
800-XXXX through 855-XXXX line numbers:
- Reserved for deaf, hard of hearing, or speech impaired
900-<2-9>XX-XXXX codes:
- Premium services billed to calling party
- Service provider IDis embedded in the <2-9>XXcode
ANI II digits:
- Two-digit pairs sent withoriginating number to identifythe type
of originating station (POTS, pay phone, and soon)
There areeight N11 codes, called service codes, which are not usedas areacodes. These are
three-digit codes in the Nl 1 format:
N11 Code Assignments
N11 Code Description
211 Community information and referral services (United States)
311
Nonemergency police and other governmental services (United
States)
411 Local directory assistance
511 Traffic and transportation information {United States); reserved
(Canada)
611 Repair service
711 Telecommunications relay services (TRS)
811 Business office
911 Emergency
In some U.S. states. Nl I codes that are not assigned nationally can be assigned locallj. if the
local assignments can be withdrawn promptly if a national assignment is made. There are no
industry guidelines for the assignment of Nil codes.
Additional NANP reserved area codes include the following:
456-<2-9>XX-XXXX numbers: These codes identify carrier-specific services by
providing carrier identification within the dialed digits. The prefix following 456. <2-
9>XX. identifies the carrier. Use of these numbers enables the proper routing of inbound
international calls, destined for these services into, and between, NANP area countries.
) 2010 Cisco Systems. Inc. Dial Plan Implementation
555-01XX linenumbers: These numbers are lictitious telephone numbers thatthe film
industry, education, various demonstrations, and others can use. Ifanyone dials one of
these numbers, it docsnot cause a nuisance toanyactual person.
800-XXXX through 855-XXXX line numbers: These numbers are in the fonnal 800-855-
XXXX and prov ide access to PSTN services for deaf, hard-of-hearing, orspeech-impaired
persons. Suchserv ices include TRSand message relaysen ice.
900-<2-9>\\-X.\XX numbers: These codes are for premium services, withthe cost of
each 900call billed tothecalling party. 900-<2-9>XX codes, each subsuming a block of
10.000 numbers, areassigned toservice providers who provide andtypically bill for
premiumservices. These service providers, in turn, assign individual numbers to their
customers.
Implemenling Cisco Voice Communications and QoS (CVOICE) vS.O 2010 Cisco Syslems, Inc
European Telephony Numbering Space
This subtopic describes the European Telephony Numbering Space (ETNS).
European Telephony Numbering Space
The ETNS is a Europeannumbering space that is parallel to
existing national numberingspaces.
It is administered byEuropean Telecommunications Office
(ETO).
There are four ETNS services available:
- Public Service Application
- Customer Service Application
Corporate Network
- Personal Numbering
European Numbering Structure
Country Code/
Group ID Code
European
Service Code
European
Subscriber Number
European Service Identification (ESI)
The ETNS isaEuropean numbering space that isparallel tothe existing national numbering
spaces and isused toprovision pan-European services. Apan-European service isan
international service that can be invoked from at least two European countries.
The European Telecommunications Office (ETO) Administrative Council supervises the
telecommunications work ofthe European Radiocommunications Office (ERO). This
supervision includes the establishment, detailing, and change ofETNS conventions and the
designation of European Service Identification (ESI) for new ETNS services.
The main objective of ETNS isto allow effective numbering for European international
sen ices forwhich national numbers maynot beadequate andglobal numbers may not be
available. The designation of a new European country code, 388, allows European international
companies, services, and individuals toobtain asingle European number for accessing their
services.
Four ETNS services are now available: Public Service Application, Customer Service
Application. Corporate Networks, and Personal Numbering. An ESI code isdesignated for each
ETNS sen ice. The one-digit code follows the European CountryCode388 and European
Sen ice Code 3 (3883). as shown in the table.
4-10
ETNS Service and ESI Codes
ETNS Service ESI
Public Service Application 3883 1
Customer Service Application 3883 3
Corporate Networks 3883 5
Personal Numbering 3883 7
Ihefigure shows the structure of a standard international number. The initial part that is known
as the ESI consists ofthe countrv code and groupidentification codethat identifies the ETNS
(3883). followed bv a European Senice Code that identifies a particular ETNS service. The
European Subscriber Numberis the numberthat is assigned to a customerinthe contextofthe
specific sen ice. The maximum length of a European Subscriber Number is 15 digits; for
example. 3883 X XXXXXXXXXX.
Implementing Cisco Voice Communications and QoS (CVOICE) v8 D 2010 Cisco Systems, Inc
Fixed and Variable-Length Numbering Plan Comparison
This subtopic explains the fixed and variable-length numbering plans and compares their
characteristics.
Fixed and Variable-Length Numbering
Plan Comparison
Examples:
Within U.S.: -9-1-200-555-1234'', "9-200-555-1234", or "9-555-1234"
U.S. to Germany: "9-011-49-30-9876543"
. Within Germany: "0-0-30-9876543" or "0-9876543" (within the same
area code)
Germany to US.: "0-00-1-200-555-1234"
Components
Demple
Country code
Area code
Subscrfcer number
Access code
International prefix
Fixed Numbering Plan
NANP
1
3 digits
3-digitexchange code
4-digit station code
9
011
Variable-Length
Numbering Plan
Germany
49
2-4 digits
5-8 digits
0
00or +
Afixed numbering plan, such as found in North America, features fixed-length area codes and
local numbers. An open numbering plan, as found in countries that have not yet standardized on
numbering plans, features variance in the length ofthe area code or the local number, or both.
Acountry code is used to reach the particular telephone system for each country or special
service.
An area code is tvpically used to route calls lo aparticular city, region, or special service.
Depending on the region, it may also be referred lo as aNumbering Plan Area, subscriber trunk
dialing code, national destination code, orrouting code.
The subscriber number represents the specific telephone number tobe dialed, but does not
include the country code, area code (ifapplicable), international prefix, ortrunk prefix.
Atrunk prefix refers to the initial digits to be dialed in adomestic call, prior to the area code
and the subscriber number.
An international prefix is the code dialed prior to an international number (the country code, the
areacode if any. andthen the subscriber number).
The table in the figure contrasts the NANP and avariable-length numbering plan (Gennanv
numbering plan in this example).
E.164Addressing
This subtopic describes the F..164 addressing standard for defining numbering plans.
E. 164 is an international numbering plan for public
telephony systems:
*Avalid number contains the following components:
Each numbercan be upto 15digits long
The E.164 planwas developed bythe ITU
International public telecommunication number for geographic areas'
1 ! 3 5 6 7 8 9 10 11 12 13 14 15
-| ' r
Counlry coda.;cci 1 Len9lh n01 defined
is 1-3 digits
Length nol defined
National (significant) number
Maximum digits 15 - cc
E. 164 is an international numbering plan for public telephone systems in which each assigned
number contains aone-, two-, orthree-digit country code (CC) that is followed by anational
destination code (NI)C) and then by asubscriber number (SN). An E.164 number can have up
to 15 digits. The UT' originally de\eloped the E.164 plan.
In the E.164 plan, each address is unique worldwide. With up to15 digits possible in anumber,
there are 100 trillion possible E.164 phone numbers. This makes itpossible, in theory, to direct-
dial from an> conventional phone set to any other conventional phone scl in the world by
dialing no more than 15single digits.
Most telephone numbers belong tothe E.164 numbering plan, although ihis does not include
internal pri\uteautomatic branch exchange (PABX) extensions.
IheE.164 numbering plan fbr telephone numbers includes the following plans:
Country calling codes
Regional numbering plans, such as the following:
ETNS
NANP
Various national numbering plans, such as the U.K. National Numbering Scheme
Scalable Numbering Plans
This topic describes the characteristics ofa scalable numbering plan.
Attributes of a Scalable Numbering Plan
* Dial plan logic distribution
Hierarchical numbering plan
Summarization
Simplicity in provisioning
> Reduction in postdial delay
Availabilityand fault tolerance
Conformance to public standards
Scalable telephony networks require well designed, hierarchical telephone numbering plans. A
hierarchical design has these five advantages:
Simplified provisioning: Ability toeasily add new numbers and modify existing numbers
Simplified routing: Keeps local calls local and uses a specialized number key, such asan
area code, for long-distance calls
Summarization: Allows the grouping of numbers in number ranges
Scalability: Leavesspace for future growth
Management: Control from a single management point
When designing a numbering plan, thesefour attributes should beconsidered toallowsmooth
implementation:
Minimal impact on existing systems
Minimal impact on users ofthe system
Minimal translation configuration
Consideration of anticipated growth
i 2010 Cisco Systems, Inc Dial Plan Implementation 4-13
Nonoverlapping Numbering Plan
This subtopic describes nonoverlapping numbering plan:?
Nonoverlapping Numbering
1001-1999
SiteA
StteB
SiteC
SileD
Oxxx. 9xxx
txxx
2xxx
4[0-4]xx
4[5-9]xx
[6-8]xxx
Reserved
Sile A extensions
Site B extensions
Site C extensions
Site D extensions
Available for future needs
User dials 1001 to
reach local endpoint
2001-2999
User dials 2001 to reach remote endpoint
A dial plan can be designed so that all extensions within the system are reached in a uniform
\va\. fhat is. a fixed quantity of digits is used to reach a given extension fromany on-net
origination point. Uniformdialing is desirable because of its simplicity. A user does not have to
remember different ways to dial a number when calling from various on-net locations. The
figure shows an example of a four-digit uniform dial plan:
Oxxx and 9xx\ ranges are excluded due to off-net access code use and operator dialing. In
such a system, where 9 and 0 are reserved codes, no other extensions can start with 0 or 9.
Site A has been assigned the range Ixxx. allowing for up to 1000 extensions.
Site B has been assigned the range 2xxx. allowing for up to 1000extensions.
Sites C and D were each assigned 500 numbers from the 4xxx range.
The ranges 6xxx. 7xxx. and 8xxx are reserved for future use.
Altera given quantity of digits has been selected and the requisite ranges have been excluded
(for example, ranges beginning with 9 or 0), the remaining dialing space has lo be divided
between all sites. Most systems require that two ranges be excluded, thus leaving eight different
possibilities for the leading digit ofthe dial range. The table in the tigure is an example ofthe
distribution of dialing space for a typical four-digit uniform dial plan.
Implementing Cisco Voice Communications and QoS (CVOICE) v8.0 ) 201C Cisco Systems, Inc
Scalable Nonoverlapping Numbering Plan Considerations
This subtopic explains considerations that are related to anonoverlapping numbering plan.
Scalable Nonoverlapping Numbering
Plan Considerations
Intrasiteand intersite calls can use the same number of digits
- Forexample, users dial 4 digits toreachanyenterprise
destination
The call is forwarded to a local recipient or to a remote
user, depending on the dialed number
Requires centralized design
- Impractical in real life
- Careful numbering design needed from the beginning
In anonoverlapping numbering plan, all extensions can be addressed using the same number of
digits, making the call routing simple and making the network easily manageable. The same
number length isused toroute the call toan internal user and a remote user.
The disadvantage ofthe nonoverlapping numbering plan is that itis often impractical in real
life. It requires acentralized numbering approach and acareful design from the very beginning.
Overlapping Numbering Plans
This topic describes overlapping and poorly structured numbering plans.
Overlapping and Poorly Struct!
Numberins
1001-1099 3986-3999
3000-3157
3365-3985
In the example, site A endpoints use director)' numbers 1001 to 1099, 3000 to 3157, and 3365
to3985. At site B. 1001 to 1099 and 3158 to3364 areimplemented. Site Cuses ranges 1001 to
1099 and3986 to 3999. There aretwo issues with these directory numbers: 1001 to 1099
o\erlap: these directon numbers exist at all three sites, sothey are not unique throughout the
complete deployment. Inaddition, the poor structure of splitting Ihe range 3000 to3999 would
require mam entries in call-routing tables because the ranges cannot be summarized by one or
a few entries.
There arethree wa\s tosolve o\ erfapping and poorly structured directory number problems:
Redesign the directon number rangesto ensurenonoverlapping. well-structured directorv
numbers.
Use an intersite access codeand a site code that will be prepended to the directory number
to create unique dialable numbers. For example. \ou could use an intersite code of 8.
assigning site A the site code 81. site R the site code 82. and site C the site code 83.
Donot assign direct inward dialing(DID)numbers: instead, publish a singlenumberand
use a receptionist or auto-attendant.
Overlapping Numbering Plan Example
This subtopic explains asample solution lo the overlapping plan.
Overlapping Numbering Plan Exampl
1001-1999
Site A
SiteB
SiteC
SiteD
Sitecocie iniefsiteprefix
1XXX
1xxx
1xw
2xxx
SileA:code11
WAN-
11
12
13
14
Site B
User dials 6-12-1001 to reach
remote endpoint
8
1001-1999
The figure illustrates the most common solution to the overlap problem in numbering plans.
The principle of site-code dialing introduces an intersite prefix (8, in this example) and asite
code (1 xin this example) that must be prepended when dialing an internal extension in another
site With this solution, asite Auser dials afour-digit number starting with 1to reach alocal
extension and enters aseven-digit number starting with 8to reach an endpoint in aremote site.
Ihe intersite prefix and the site code that is used in this scenario show sample values and can
be set differently according to enterprise requirements. For example, the intersite prefix is
commonly set to 8and Ihe access code to 9in an NANP region, while the intersite prefix is
typically 9and the access code 0in Europe.
Scalable Overlapping Numbering Plan Considerations
This subtopic provides considerations that relate to Ihe site-code dialing solution ofoverlapping
numbering plans. Kt h
Scalable Overlappini
Intersite calls use additional intersite prefix and site code
Example:
- Intrasite: 4 digits
- Intersite 7digits (1-digit intersite prefix +2digits site
code +4 digits internal number)
: Decentralized design
Common in real life
Less preplanning necessary
Intersite prefix should cause no ambiguity with internal
numbers
Forexample. If intersite prefix is 8, internal numbers
should not start with 8
The site-code dialing solution ofthe overlap issue in numbering plans is useful in real life, as it
allows adecentralized approach to the numbering effort. Kvcn various departments within an
organization can manage their own addressing space, and the site codes can interconnect them
into amanageable unified communications network. Site code dialing does not require acareful
design from the beginning and can be implemented as the enterprise grows.
The internal extensions musl not start with the intersite prefix (for example, 8) because such
entries would cause ambiguity in the dial plan. The intersite prefix notifies the call routing
de\ ice that the call is destined to aremote location and therefore should not overlap with anv
internal number.
4-18 Implemenling Cisco Voice Communicalions andQoS (CVOICE) v8.0
mm
Private and Public Numbering Plan Integration
Thistopic describes howto integrate private andpublic numbering plans.
Private and Public Numbering Plan
Example
,. _, Site Intersite
Local Range code prefix
SiteA Ixxx
SiteB 1xxx
SiteC 1xxx
Site D 2xxx
1001-1999
PSTN DIDrange
200-555-lxxx
300-555-3xxx
400-555-1234
500-555-22xx
1001-1999
User dials user dials 9-600-555-6666
1001 to reach 1oreach a PSTNendpoint
local endpoint
Called party number
transformed lo 1001
The figure illustrates an enterprise with four locations in the NANPregion. Site-code dialing
has beendesigned to allowcalls between the enterprise locations. Each site has a trunk
connection to the PSTN, with the PSTN DIDrange provided by the telephone company (telco)
operator. SitesA and Bhave DIDrangesthat allowpublicaddressing of each internal
extension. Site C has a single DIDnumber with an interactive voice response (IVR) solution
that prompts the callers for the number ofthe internal extension for forwarding inbound calls to
the intended callee. The DIDrange of site D covers some internal extensions and must be
combined with an IVR to provide inbound connectivity to others.
Access code 9 identifies a call that is destined to an external PSTN recipient. In this example,
internal users dial 9-600-555-6666 to reach ihe PSTN endpoint.
Here are some ofthe challenges that you face with numbering plan integration:
Varying number lengths: Within the IP network, consideration is given to varying
number lengths that exist outside the IP network. Local, long-distance, and international
dialing from within the IP network may require digit manipulation.
Necessity of prefixes or area codes: It can be necessary lo strip or add area codes, or
prepend or replace prefixes. Reroutingcalls fromthe IP network to the PSTN for failure
recovery can require extra digits.
>2010 Cisco Systems, Inc. Dial Plan Implementation
Private and Public Numbering Plan Integration Functions
This subtopic explains the functions that arc provided by the integrated privateand public
numbering plans.
4-20
Private and Public Numbering PI
Integration Functions
Each siteieaches ihe
PSTN via lis loca
gateway
No DID, aulo-attendant used
Site 0:400-555-1234
Primary Path
1001-1999
The three basic features that are provided by the integrated private and public numbering plans
include the following:
Reachability to external PSTN destinations: Internal users get access to external
destinations over the gateway, which acts like a junction between the private and public
addressing scheme.
Auto-attendant: IVR system Is required to provide connectivity to internal extensions
when a sufficient DID range is not available.
PSTN acts a backup path in case the IP WAN fails or becomes congested: In such
cases, the gateways redirect the intersite calls over the PSTN to provide uninterrupted
senice.
Private and Public Numbering Plan Integration Considerations
This subtopic provides considerations that are related tothe private and public numbering
integration.
Private and Public Numbering Plan
Integration Considerations
Access code and intersite prefix must not cause ambiguity with
internal numbers
-- Access code example; 9
Internal numbers should not start with 9
- Intersite (WAN) prefixexample: 8
Internal numbersshould notstartwith 8
- Path selection transparentto user
- User dials WAN (intersite) prefix
- Call may go through WAN or PSTN
* Auto attendantfbr non-DID numbers
Numberadjustment
- Prepending and removing of prefixes or access code
PSTN backupexample; replace 8-12with 300-555
- Mapping ranges
Example: Internal 1xxx to DID 300-555-3xxx
Whenintegrating privateandpublicnumbering plans, special consideration shouldbe givento
these aspects:
Noambiguity with the internal and intersitedialing: The prepended access codeshould
uniquely identifyall calls that shouldbreakout to the PSTN.
Path selectiontransparent to the user: Usersdial site codes to reach remotelocations and
the intersite calls select the IP network as the primary path. If the IP WANis unavailable,
the call should be redirected over the PSTN. The user does not need to take any action for
the secondary path to be chosen.
Auto attendant for non-DID numbers: When the DID range does not cover all internal
extensions, an auto-attendant is needed lo allow inbound calls.
Number adjustment: The voice gateway needs to adjust the calling and called numbers
when a call is set up between the sites or via the PSTN. One manipulation requirement
arises when an intersite call is rerouted over the PSTN, 'fhe intersite prefix and site code
(for example8-12)must be then replacedwitha publicnumberidentifying the location, for
example 300-555. Another type of manipulation is needed to map the internal ranges to
DID scopes, for example. Ixxx to 300-555-3xxx.
Number Plan Implementation Overview
This topic pro\ides an overviewof number plan implementation.
Private Number Plan Implementation
* Select length of internal number and site code
Based on number of users per site, number of sites
Local endpoints reachable via automatically created dial-
peers for registered directory numbers
Remote endpoints reachable via configured VoIP dial peers
* Number ambiguity breaks number plan
WAN.
The implementation ofthe private numbering plan takes into account the number of users per
site and the number of sites. The length ofthe internal numbers and the site codes must match
the size ofthe environment and at the same time allow space for future growlh. The figure
illustrates that the internal extensions can consist of two to four digits, and the site codes can
consist of one to three digits.
Call routing to local endpoints is achieved automatically because the registering endpoints have
virtual dial peers that are associated with them. Thedial peers ensure that calls are routed to the
registered phones based on their directory numbers.
Call routing to remote locations is enabled b> VoIP dial peers that describe the primary path
over the IP WAN.
The dial plan must not be ambiguous to ensure correct call routing.
Private Number Plan Implementation Example
This subtopic provides aprivate number plan implementation example.
Private Number Plan Implementation
Example
One large site, 10+ small sites
1001-7999
101-799 SiteD:code24
dial -peer voice 1 vo
P
des tination-pattern 821.
BBS aion target ipv4 10.1 1 1
di: -peer voice 2 vo ip
seasion target ipv4 10.2 2 2
dial -peer voice 3 vo
^-P
des tiaotion-pa -tern 82 3.
BOB
aion target ipv4 10.3 3 J
Site A: code 21
10.1.1.1
IPWAW
101-799
-799
In the example inthe figure, the enterprise has one large site (site A)with 7000 users and
several smaller sites with less than 700 users each. The codes for all sites are two-digit numbers
(21 to40). The internal extensions in the large site are four digits long (1001-7999), while the
extensions inthesmaller sitesarethree digitslong (101 lo799). To implement thedial plan,
VoIP dial peers are configured with destination patterns that match seven-digil numbers in the
large site and six-digit numbers inthe remaining sites, starting with theintersite prefix 8.
Public Number Plan Implementation
This subtopic describes how toimplement the public numbering plan.
Public Number Plan impi
Public numberingplan imposed by telco operator
Enterprise may choose the size of DID range
May have financial impact
Gateways can map number ranges
Decouples internal frompublic numbering
For example: 200-555-3xxx <->1xxx
Complex mapping scenarios difficult to implement
For example: 200-555-3xxx <-> 1xxx + 50
Should be avoided during design
The enterprise docsnot design its public numbering plan. It is imposed by thetelco operator,
fhe enterprise mav influence the si/e ofthe DIDrange, which is often related to a financial
decision.
Gatewav s prov idea mapping bctweenthe DI Dandtheinternal number ranges. Forexample,
the PSTNDIDrange 200-555-3xx.\ can be easily converted to lxxx and back when calls
traverse the gatewav. Complex mapping formulas (lor example, mapping of 200-555-3xxx to
lx\x + 50) are too complex lo implement and should be avoided.
4-24 Implementing Cisco Voice Communicalions and QoS (CVOICEj v8 0
Call Routing Overview
Thistopic provides anoverview of call routing.
Call Routing Refresher
HH IProutina | Call routing mmm
Static/dynamic Only static.
IP routing table Dial plan.
IP route Dial peer.
Hop-by-hop routing
each router makes an
inde pe nd ent d ecision
Inbound and outbound call legs. The gateway
negotiates VoIP parameters with preceding and
next gateways before a call is forwarded.
Destination-based routing
Called number, matched by destination-pattern, is
one of many selection criteria.
Most explicit match rule The most explicit match rule fbr destination-pattern
exists butaiso other criteria considered.
Equal paths Preference can be applied to equal dial peers. Ifail
criteria are the same, random selection.
Default route
Possible. Often points at external gateway or
gatekeeper.
Themost relevant properties of call routing canbecompared to thecharacteristics of IPpacket
routing.
fhe entries that define where to forward calls are the dial peers. All dial peers together build
thedial plan, which isequivalent tothe IP routing table. Thedial peers arestatic innature.
Hop-by-hop call routing builds ontheprinciple of call legs. Before a call routing decision is
made, thegateway must identify theinbound dial peerandprocess itsparameters. Thisprocess
may involve VoIP parameter negotiation.
The call routing decision is the selection ofthe outbound dial peer. This selection is commonly
basedon the callednumber, whenthe destination-pattern command is used. The selection
may bebased onother information, and that other criteria may have higher precedence than ihe
called number. When the called number is matched to find the outbound dial peer, the longest
match rule applies.
If more than onedial peerequally matches thedial string, all ofthe matching dial peersare
used to forma so-called rotary group. The router attempts to place the outbound call leg using
all ofthe dial peersintherotary group untilone is successful. Theselection orderwithin the
groupcan be influenced by configuring the preference.
Adefault call route canbe configured usingspecial characters when matching the number.
Call Routing Example
Thissubtopic provides a call routingexample.
-.1 (*!
bbLtfaJI
!> * 300 3U 2in
Dial 8-22-1001
1001
Primary Path
Call progressed to 1001 in site 22
Onginating gateway strips 8-22
IP WAN
PSTN
. AX :Kwu<
Secondary Palh (Used When WANUnavailable!
Call progressed to 300-555-3001
Digit manipulationrequired on onginatingand terminatinggateways
Ihc voice gateways inthis example are faced with the task of selecting the best path for a given
destination number. Such a requirement arises when thepreferred pathgoesthrough the IP
WAN. The backup PSTN path should be chosen when the IP WAN is either unavailable or
lacks the needed bandwidth resources.
Ihc figure illustrates a scenario with two locations that are connected lo the IP WAN and
PSTN. When thecall goesthrough the PSTN, its numbers (both calling andcalled) have tobe
manipulated so that they are reachable within the PS'I'N network. Otherwise, the PSTN
switches will not recognize the called number andthecall will fail. Digit manipulation is
closely related to call routing.
2010 Cisco Systems Inc
Summary
Summary
Anumbering plan defines how toallocate numbers and can
use eitherfixed (such as NANP) or variable (suchas ETNS)
number lengths.
Ascalable numberingplan should be hierarchical, simpleto
summarizeand provision, and conform to standards.
- Overlappingnumbering plans require that site codes are
used to reach remote endpoints.
Private and public numbering planintegration defines the
mapping between internal and DID rangesand affects the
necessary digit manipulation.
- An internal numbering plan corresponds to the telephone
numbers assigned to internal endpoints.
Voice gateways route calls using the longest-match principle.
4-28 ImplementingCisco Voice Communicationsand QoS (CVOICE) v8.0 2010 Cisco Systems, Inc
'mm
*_
Lesson 2
Understanding Dial Plans
Overview
Adial plan isthe central part ofany telephony solution and defines how calls are routed and
interconnected. Adial plan consists of various components, which can beused invarious
combinations. Thislesson describes the components of a dial planandhowthey are used on
Cisco IOS gateways.
Objectives
Upon completing this lesson, you will be able todescribe the components ofadial plan and
their functions. This abilityincludes beingable to meet theseobjectives:
Describe the characteristics and components of a typicaldial plan
Explain the concept ofendpoint addressing, including overlapping directory numbers
Describe the characteristics of call routingandthe importance of path selection
Explain PSTNdial plan requirements
Describe special ISDN dial plan requirements
Provide thecharacteristics of digit manipulation ina voice gateway implementation
Describe the implementation of calling privileges in a voice gateway
Describe the characteristics of call coverage in gateway implementation
Defining Dial Plans
fhis topic prov ides an overview ofthecomponents and requirements ofadial plan.
Dial Plan Componei
Adial plan defines how to manage calls:
Endpoint addressing: Internal destination accessibility can be
provided by assigning directory numbers to all endpoints.
* Call routing and path selection: Different paths can be
selected to reach the same destination.
Digit manipulation: Digits can be manipulated priorto or after
a routing decision has been made.
* Calling privileges: Different groups of devices can be
assigned to different classes of service by granting or
denying access to certain destinations or resources.
1 Call coverage: Special groups of devices can be created to
manage incoming calls for a certain service according to
different rules, avoiding dropped calls.
Although most peopleare not acquainted withdial plans b> name, they use themdaily. Adial
plandescribes the process of determining howmanyand which digits are necessary for call
routing, if the dialeddigitsmatchthe numberand patterns, thecall is processed and fonvarded.
Designing dial plansrequiresknowledge ofthe network topology, dialingpatterns, and traffic
routing requirements, fhere are no dynamic routingprotocols for E.164 telephony addresses.
VoIP dial plans are statically configured on gateway and gatekeeper platforms.
A dial plan consists of these components:
Kndpoint addressing (numbering plan): Assigning directory numbers to all endpoints
and applications (such as voice-mail systems, auto attendants, and conferencing systems)
enables vou to access internal and externa! destinations.
Call routing and path selection: Multipledifferent paths can lead to tlie same destination.
Asecondary pathcanbe selected \\ henthe primary pathis not available. Forexample, a
call can be transparently rerouted over the public switched telephone network (PSTN)
during an IP WAN failure.
Digit manipulation: Manipulation ofthe numbers before routinga call, for example, when
a call is rerouted over the PSTN, This can occur before or aller the routing decision.
Calling privileges: Different privileges canbe assigned to variousdevices, grantingor
deny ing accessto certaindestinations. For example, lobbyphones may reachonly internal
destinations, while executive phones could have unrestricted PS'I'N access.
Call coverage: You can create special groups of devices to manage incoming calls for a
certain service according to different rules (lop-down, circular hunt, longest idle, or
broadcast). This also ensures that calls are not dropped without being answered.
#*
This subtopic describes the components ofadial plan and how they arc implemented on Cisco
voice gateways.
Dtat Plan Component
Endpoint addressing
Call routingand path selection Dial peers
Digit manipulation
Calling privileges Class of restriction (COR) names and lists
Call coverage
Cisco IOS Gateway
POTSdial peers for FXSports, ephone-dn
and voice register directory number
voice translation profile, prefix, digit-strip,
forward-digits, and num-exp
Call hunt, hunt groups, call pickup, call
waiting, call forwarding, overlaid directory
numbers
FXS = Foreign Exchange Station
CiscoIOS gateways, including Cisco Unified Communications Manager Express andCisco
Unified Survivable Remote Site Telephony (SRST), support all dial plan components.
Thetableprovides anoverview of themethods that CiscoIOS gateways useto implement dial
plans.
Dial Plan Requirements
This subtopic explains the common requirements of a dial plan.
Dial Plan Requirements
Site A:
Site Code. 21
DID" 200-555-2XXX
Site B:
Site Code: 22
DID 300-555-3XXX
1001 1002
Dialing from site example
1002 (local userj
8-22-1001 (user in other site)
9-400-555-444A (PSTN phone)
PSTN
400 55S 4444
Dialing from PSTN example
1-200-555-2001 (user in site A)
1-300-555-3001 (user in site B)
The figure shows a typical dial plan scenario. Calls can either be routed via an IP WAN link or
a PSTN link, and routing should work for inbound and outbound PSTN calls, intrasitc calls,
and intersite calls.
The dial plan defines the rules that govern how a user reaches any destination. Definitions
include the following:
Extension dialing: Determines how many digits must be dialed to reach an extension
Extension addressing: Determines how many digits are used to identify extensions
Dialing privileges: Allows or disallows certain types of calls
Path selection: Selects one path from several parallel paths
Automated selection of alternate paths in case of network congestion: 1or example,
using the local carrier for international calls iflhe preferred international earner is
unavailable
Blocking of certain numbers: Prevents unwarranted high-cost calls
Transformation ofthe called party number
Transformation ofthe calling party number
A dial plan suitable for an IP telephony system is not fundamentally different from a dial plan
that is designed for a traditional telephony system. However, an IP-based system presents
additional possibilities. In an IP env ironment. telephony users in separate sites can be included
in one unified IP-based system. These possibilities by IP-based systems require you to think
about dial plans in new ways. This lesson examines the elements that you must consider to
design the dial plan.
Endpoint Addressing
This topic describes endpoint addressing.
Endpoint Addressing
Internal endpoints become reachable by assigning directory
numbers.
Directory numbers are assignedtoendpoints (phones, fax
machines, etc.) and applications {voice-mailsystems, auto-
attendant, etc.).
The number of extensions required generallydetermines Ihe length
of directory number digits.
DID numbers for inbound PSTN calls are mapped to internal
directory numbers.
Cisco Unified
Communicalions
Manager Express
Phone Numbers
Assigned to Endpoints
1001 1002 1003
Cisco Unity
Express
1099 8001
Reachability ofinternal destinations isprovided by assigning directory numbers toall
endpoints (such as IP phones, fax machines, and analog phones) and applications (such as
voice-mail systems, auto-attendants, andconferencing systems).
Thenumber ofdialable extensions determines the quantity of digits thatareneeded todial
extensions. For example, a four-digit abbreviated dial plan cannot accommodate more than
10.000 extensions (from 0000 to9999). If0and 9are reserved asoperator code and external
access code, respectively, the number range isfurther reduced to8000 (1000 to 8999).
Ifdirect inward dialing (DID) isenabled for PSTN calls, the DID numbers are mapped to
internal directory numbers.
Dial Plan Implementalion 4-33
Endpoint Addressing Considerations
This subtopic explains endpoint addressing considerations.
Endpoint Addressing Considerations
Directory numbersare assigned to endpoints.
* Internal extensions are mapped to inbound PSTN calls.
Often dependent on range of DID numbers
Auto-attendant can be used for non-DID numbers
* The biggest challenge: Creating an endpoint addressing
scheme in multisite environments
Primarily a CiscoUnified Communications Manager
Expressor Cisco Unified Communications Managerissue
Themost common issue with endpoint addressing is related tothemapping of internal
extensions to available DID ranges assigned by the PSfN. When the DID range covers the
entire internal address scope, an auto-attendant can be used to route calls between the PS'I'N
and the internal network.
One ofthe biggest challenges when creating anendpoint addressing scheme for a multisite
installation istodesign a flexible and scalable dial plan thai has noimpact onthe enduser. The
existing ov erlapping directory numbers present a typical issuethat must be addressed.
Endpoint addressing is primarily managed by the call agent, suchas Cisco Unified
Communications Manager or Cisco Unified Communications Manager Hxpress.
4-34 Implementing Cisco Voice Communicalions and QoS (CVOICE] v8 0
Call Routing and Path Selection
This topic describes call routing and path selection.
Route the call depending on the dialed number.
Select the best path.
- Which deviceto send the call to?
- Backup path available if first choice not available?
Handled by dial peers on Cisco IOS gateways:
- Inbound and outbound dial-peer matching determines the
routing.
Call routing and path selection are the dial-plan components that define where and how calls
should be routedor interconnected. Call routing usually depends on the called number- ht is.
destination-based eall routing is usually performed. This is very similar to IP routing wh, h
also relies on destination-based routing. Multiple paths to the same destmation might ex. t,
especial* in multisite environments-for example, apath using an IP connection or apath
using aPSTN connection. Path selection helps you decide which ofthe available paths should
be used.
Avoice gateway might be involved with call routing and path selection depending on the
protocol and design that is used. For example, an H.323 gateway will at least mute: the call
between the call leg that points to the eall handler and the call leg that points to the PS IN.
When aCisco IOS gateway performs call routing and path selection, the key components that
are used are dial peers.
Path Selection Example
fhis subtopic presents an example ofpath selection.
Path Selection Example
In the example, to call 300-555-2001, use:
IP WAN as the preferred path
..' If WAN not available, secondary path through PSTN
IP WAN
PSTN
300-555-2001
1001
In the ligure. if auser dials an extension number in another location {8-22-2001) the eall
should be sen, ov er the IP WAN, If the WAN path is unavailable (network failure, insufficient
bandwidth, or no response, the call should use the local PSTN gateway as abackup and send
the call through the PSTN.
For PSTN-routed calls, digit manipulation must be configured on the gatewav to transform the
internal numbers to1,164 numbers that can bedialed inthe PSTN.
4-36 Implementing Cisco Voice Communications and QoS (CVOICE) vS.O
2010 Cisco Systems, Ir
PSTN Dial Plan Requirements
Thistopic describes PSTN dial planrequirements.
PSTN Dial Plan Requirements
PSTN
Requirements
Inbound call
routing
Outbound call
routing
Correct calling
party number
presentation
Dial Pian Components
Call routing and path selectionfor inbound PSTN dial
peer to outboundVoIPor localdial peer
Digitmanipulation to transform inbound called party
number to endpoints
Call routing and path selectionfor inbound VoIPor
local dial peer to outbound PSTN dial peer
Digit manipulation Id transform outbound called party
number to PSTN requirements
Digitmanipulation to transform calling party number
to meet PSTN requirements
A PSTNdial plan has three key requirements:
Inbound call routing: Incoming calls from the PSTN mustberouted correctly totheir
final destination, which mightbe a directly attached phone or endpoints that are managed
byCisco Unified Communications Manager or CiscoUnified Communications Manager
Express. This inbound call routing also includes digit manipulation toensure thatthe
incoming called number matches thepattern that is expected bythe final destination.
Outbound eall routing: Outgoing callstothe PSTN mustberouted tothevoice interfaces
ofthe gatewayfor example, a Tl/El or a Foreign Exchange Office (FXO) connection. As
withinboundcalls, outbound calls may also require digit manipulation to modify the called
number according tothe PSTN requirements. Thisoutbound call routing usually includes
stripping of anyPSTN access codethatmight beincluded in theoriginal called number.
Correct PSTN calling party number presentation: An often-neglected aspect is the
correct calling number presentation forbothinbound andoutbound PSTN calls.The calling
number for inbound PSTN calls is often left untouched, which may have a negative impact
on theend-user experience. Thecalling number that is presented to theendusershould
include the PSTN access code and any other identifiers that are required by the PSTNto
successfully place a call usingthat calling numberfor example, using themissed calls
directorv.
Inbound PSTN Calls
This subtopic describes hou gateways manage inbound PSTNcalls.
inbound PSTN Cal
Gateway modifies called
number lo 1001 and
routes lo IP phone
PSTN
Call setup from PSTN
Called numbet
200-555-2001
User dials 1-200-555-2001
Unified CM Eiptess - Cisco Unified Communications Manager Express |
The example illustrates hou gatev\avs manage inbound PSTN calls. The site consists ol'a Cisco
UnifiedCommunications Manager Express systemwith endpoints registered to it, connected to
the PSTNover a digital trunk. The DIDrange ofthe PSTNtrunk is 2005552XXX. and phones
usethe extension range IXXX. The processing of an inbound PSTN call occurs inthesesteps:
Stepl
Step 2
Step 3
Step 4
A PSTN user places a call to 1-200-555-200I-
evtension 1001,
-that is, an endpoint with internal
The call setup message is received bv the gateway with a called number of 200-555-
2001.
The gateway modifies the called number lo 1001 and routes the call to the voice port
that v\as created when the Cisco Unified IP phone registered with Cisco Unified
Communicalions Manager Express.
flic phone rings.
Implemenling Cisco Voice Communications and QoS (CVOICE) v8 0 2010 Cisco Syslems. Inc
Numbers in Inbound PSTN Calls
This subtopic explains how gateways must transform called and calling numbers when
forwarding inboundPSTN calls.
Numbers in Inbound PSTN Calls
Called party number converted from public PSTN format to
internal extension to match the outbound (virtual) dial peer
Access code and long-distance prefixprepended to calling
number to allow callback
1001 DID: 200-555-2XXX
Called partynumber 200-555-2001 1001
Calling partynumber 300-555-3002 9-1-300-555-3002
Thefigure provides a comprehensive description ofthe required number manipulation when a
gateway receives an inbound PSTN call. Both the called and calling numbers must be
transformed:
The called number must be converted fromthe public E.164 format to the internal number
used for internal dialing. Thistransformation ensures thatthecall matches theoutbound
dial peer that isautomatically created at endpoint registration (thedirectory numbers are
commonly configured withtheir internal extensions).
Thecalling number must be presented to thecallee ina way that allows callback. Because
access codesare commonly usedto reachexternal destinations, the callingnumber
forwardedto the destination should include the access code. Optionally, some region-
specific prefixes may have tobeadded, such asthelong-distance prefix inthe North
American Numbering Plan (NANP) region, "1."
Outbound PSTN Calls
fhis subtopic describes how gatewavs manage outbound PSTN calls.
User dials
9-1-300-555-6001
Gaieway modifies calling ar*f called party nu
Calling 1001 -> 2005552001
Called 913005556001 - 13005556001
Q 931 call selup
Called number 1-300-555-6001
Calling number 200-555-2001
"unilied CM Express - Cisco Unified Communicalions Manager Express I
The figure shows the call How for an outbound call. The site consists of a Cisco I inified
Communications Manager Express s\ stemwithendpoints registered to il. connected to the
PSTNover a digital trunk. The access code is 9. T'he processing of an outbound PSTNcall
occurs in these steps:
Step 1 I !ser places a call to 9-1-300-555-6001 from the phone with extension 1001.
Step 2 The gatewav accepts the call and modifies the called number to 1-300-555-6001.
stripping off the PSTN accesscode 9. The gateway also modifies the callingnumber
lo 200-555-2001 by prefixing the area code and local code and mapping the four-
digit extension to the 1)11) range.
Step 3 The gatewav sends out a call setup message with the called number set lo 1-300-
555-6001 and the calling number sel to 200-555-2001.
Step 4 The PSI N subscriber telephone at 300-555-6001 rings.
4-40 Implementing Cisco Voice Communications and QoS (CVOICE| v8 0 2010 Cisco Systems. Inc.
Numbers in Outbound PSTN Calls
This subtopic explains how gateways must transform called and calling numbers when
processing outbound PSTN calls.
Numbers in Outbound PSTN Calfs
Called party number:
- Strip access code
- Prepend long-distance prefix (optional)
Calling party number:
- Convert internal extension to public PSTN format
For telco compliance and callback
DO: 200-555-2XXX
Incomin
Called partynumber 9-1-300-555-3002 1-300-555-3002
Calling party number 1001 200-555-2001
The figure summarizes the requirements for numbermanipulation whena gateway processes an
outbound PSTNcall. Both the called and calling numbers must be transformed as follows:
The called number processing involves the stripping ofthe access code. Optionally, some
region-specific prefixes mayhave to be added, suchas the long-distance prefix inthe
NANP region, "1."
The callingnumbermust be converted from the internal extensionto the public E.164
format. If the outgoingcalling number is not configured on the gateway, the telco operator
sets the value to the subscriber number, but this setting may be inaccurate if a DID range is
available. For example, the calling number for a call originating from 1002would be set to
222-555-2000. Setting the calling number is considered a good practice and ensures proper
callback functionality.
>2010 Cisco Systems, Inc Dial Plan Implementation 4-41
PSTN Backup
This subtopic describes PSTN backup of IPWAN connectivity.
PSTN Backup
If the IP WAN fails, intersite calls are rerouted over the PSTN
Two dial peers with 8(site-code).XXXXdestination pattern per site
Primary VoIP dial peer
Secondary POTS dial peer that manipulates the digits
Site A
Site Code 21
DID: 200-555-2XXX
--PSTN
Site A
Site Code: 22
DID 300-555-3xxx
1001-1999
The figure shows a multisite deployment with each site running its own Cisco Unified
Communications Manager Fxpress system. T'heprimary path for intersite calls is the IP WAN.
Por uninterrupted operations, the IP WAN should be backed up by an independent transmission
media. Both sites have access to the PSfN. therefore the PSTN offers Ihe perfect solution as a
secondary path for intersite calls.
4-42 Implementing Cisco Voice Communications and OoS (CVOICE) v8 i >2010Cisco Systems, Inc.
ISDN Dial Plan Requirements
This topicdescribes ISDN dial plan requirements.
Type of Number
TONin ISDNprovides informationabout number format:
* Subscriber
- 7-digit subscriber number
3-digit exchange code
National ,
10-digit number
3-digit area code
3-digtt exchange code
International
Variable length (11 digits for U.S. numbers)
Country code (1 digit tor U.S. country code 1)
Area code (3 digits for U.S. area code)
3-digit exchange code
Number transformations
depend on TON
The Typeot'Numbcr (TON)or Natureof Address Indicator (NAI)parameterindicates the
scopeofthe address value, such as whether it is an international number(including the countrv
code) a "national" or domestic number (without country code), and other formats such as
"local" format (without an area code). It is relevant for E.164(regular telephone) numbers.
The TON is carried in ISDN-based environments. Voice gateways must consider the TON
when transforming the called and calling numbers for ISDN calls.
2010 Cisco Systems, Inc. Dial Plan Implementation 4-43
ISDN Dial Plan Requirements
4-44
This subtopic explains which dial plan components areused to implement ISDN dial plan
requirements.
Ian
ISDN Requirements
Correct PSTN inbound ANI
presentation, depending on TON
Couect ISDN numbering plan
and TON presentation
Digit manipulation to transform
inbound PSTN ANI according to
TON
Manipulate ISDN numbering plan or
TON to meet PSTN requirements
ISDNnetworks impose new number manipulation needs, in addition to the typical requirements
for PSTN calls:
Note The calling party number in ISDN is called Automatic Number Identification (ANI). The called
party number in ISDN is referred to as Dialed Number Identification Service (DNIS).
Correct PSTN inbound AM presentation, depending on TON: Some ISDN nctworLs
present the inbound ANI as the shortest dialable number combined with the TON. This
treatment ofthe ANI can be a potential problem because simply prefixing the PSTN access
code might not result in an ANI that can be called back. The potential problem can be
solved bv proper digit manipulation on gateways.
Correct PSTN outbound AM presentalion, depending on TON: Some ISDN networks
and PBXs might expect a certain numbering plan and TON for both DNISand ANI. Using
incorrect flags may result in incompletecalls or an incorrect DNISand ANI presentation.
Digit manipulation can be used to solve these issues.
inbound ISDN Calls
This subtopic provides an example of how gateways manage inbound ISDN calls.
Inbound ISDN Calls
DID: 200-555-2XXX
1001-1999
1 Subscriber
2 National
3 Internationa!
PSTN
555-1111
400-556-2222
49-30-1234567
Site 1: 200-555-1111
Requi-ed ANI Transformation
9-555-1111
9-1-400-555-2222
9-011-49-30-1234567
In the tigure. three different calls are received at the voice gateway. The first call is received
from the local area with a subscriber TON and aseven-digit number. This number only needs
to be prefixed with access code 9. The second call, received with national TON and 10 digits, is
modified by adding access code 9and the long-distance number 1, all of which are required for
placing calls back to the source ofthe call. The third call is received from oversees with an
international TON. For this call, the access code 9and 011 must beadded tothe received
number, as a prefixto the country code.
Digit Manipulation
This topic provides an overview ofdigit manipulation and its deployment.
Digit Manipulation
* Related to call routing and path selection
Inbound calls:
Called number needs to match internally used patterns.
Calling numbershould be presented as a dialable
number.
Outbound calls:
Called number must satisfy internal or PSTN
requirements.
Calling number needs to be dialable.
Digit manipulation iscloselv related tocall routing and path selection.
Digit manipulation isperfomied for inbound calls loachieve these goals:
Adjust the called partv number tomatch internally used patterns.
Present the calling partv number as a dialable number.
Digit manipulation is implemented for outbound calls toensure the following:
Called number satisfies theinternal or PSTN requirements.
Calling number is dialable and provides callback ifsuftlcient PSTNDID ts available.
Digit manipulation is covered in a later lesson.
Calling Privileges
This topic provides an overview of calling privileges.
Calling Privileges
Defines the destinations that a user is allowed tocall
- For example, manager cannot be reached from lobby
phone
. Oftenused to control telephony charges:
- Blockscostlyservice numbers
- Restricts international calls
* Often called "class of service"
- Not the same as Layer 2 CoS
Calling privileges are equivalent to firewalls in networking. They define call permissions by
specifving which users can dial given destinations. The two most common areas of deploying
callingprivileges are as follows:
Policy-defined rules to reach special endpoints. For example, manager extensions cannot be
reached from a lobby phone.
Billing-related rules that are deployed to control telephony charges. Common examples
include the blocking ofcostly service numbers and restricting international calls.
Calling privileges are referred to as a"Class of Service," but should not be confused with the
Layer 2class of service (CoS) that describes quality of service (QoS) treatment of traffic on
Lavcr 2 switches.
Calling Privileges Example
This subtopic prov ides an example ofcalling privileges.
Calling Privileges ExampI
Site 1.
200-555-1111
DID 200-555-2XXX
PSTN
Executive Employee Lobby
Call'Permission
Executive Site 1(local), site 2(long distance), site 3(international)
Employee Site1(local), sile2 (long distance}
Lobby Site 1 (local)
The figure illustrates the tvpical deployment ofcalling privileges. The internal endpoints are
classified imo three roles: executive, employee, and lobby, liach role has a set ofdialable PSTN
destinations that are associated with it. The executives can dial any PSTN number, the
employees are allowed to dial any external numbers except international destinations, and the
lobby phones permit the dialing of local numbers only,
Ihe deployment of calling privileges is covered ina later lesson.
4-48 Implementing Cisco Voice Communications and QoS(CVOICE] v8.0
Call Coverage
This topic describes call coverage and explains why it is used inenterprise environments.
Call Coverage
Call coverage ensures that all incoming calls are
answered:
For individuals:
- Call waiting when the line is busy
- Call forwarding ifthe line is busy or not answering
For user groups:
- Call distribution
- Call pickup
Call coverage features are usedto ensurethat all incoming calls to Cisco Unified
Communications Manager Express are answered by someone, regardless of whether the called
number is busy or does not answer.
Call coverage can be deployed for two different scopes:
Individual users: Features such as call waiting and call forwarding increase the chance of
a call being answered by giving it another chance for a connection if the dialed user cannot
manage the call.
User groups: Features such as Call Pickup, call hunt, hunt groups, and overlaid directory
numbers provide different ways to distribute the incoming calls to multiple numbers and
have them answered by available endpoints.
) 2010 Cisco Systems. Inc. Dial Plan Implementation 4-49
Call Coverage Features
This subtopic provides an overview of call coverage features.
CallC
Call
forwarding
overage Features
Directory number configured to forward calls to another number
when it is busy or does not a nswet
Call hunt Multiple directory numbers with the same telephone number.
Preference and hunt stop ensure a defined call delivery order.
Call pickup Ringing calls can be picked up by another endpoint with
appropriate permissions.
Call waiting Call-waiting beep and displayed calling number inform a bout an
incoming call during active conversation. The new call can be
answered or forwarded.
B-ACD Basic Automatic Call Distribution answers calls and guides
through a selection menu.
Hunt groups Multiple extensions belong to a hunt group identified by a pilot
number. Calls to the pilot number sent sequentially to members.
Overlaid
ephone-dn
Multiple directory numbers assigned to one phone button. User
can answer calls to any number.
Cisco voice gateways provide various call coverage features:
Call forwarding: Calls are automatically diverted to a designated number on busy, no
answer, all calls, or only during night-service hours.
Call hunt: System automatically searches for an available directory number from a
matching group of directory numbers until the call is answered or the hunt is stopped.
Call Pickup: Calls to unstaffed phones can be answered by other phone users using a
softkey or by dialing a short code.
Call waiting: Calls lo busy numbers are presented to phone users, giving them the option
to answer or let them be forwarded.
Basic automatic call distribution (B-ACD): Calls to a pilot number are automatically
answered by an interactive application thai presents callers with a menu of choices before
sending them to a queue for a hunt group.
Hunt groups: Calls are forwarded through a pool of agents until answered or sent to a final
number.
Overlaid ephone-dn: Calls to several numbers can be answered by a single agent or
multiple agents.
Summary
Summary
Adial plan involves endpoint addressing, call routing and path
selection, digit manipulation, calling privileges, and call coverage.
. Thebiggestchallengeofendpointaddressingistodesigna
scalable addressing scheme that maps internal numbers toDID
numbers.
. Cisco Unified Communications Manager Express routes calls by
selecting the best match.
PSTN calls require the manipulation of calling and called numbers
when calls traverse the voice gateway.
ISDN calls require the conect TON presentation and manipulation of
thecalling and called numbers.
Digit manipulation fulfills internal and PSTN requirements and
presents thenumbers in a dialable form.
Calling privileges enforce dial permissions to control dial behavior
and prevent toll fraud.
Call coverage ensuresthatcallsare answered.
4-52 Implementing Cisco Voice Communicalions and QoS (CVOICE) v8.0 2010 Cisco Systems, Inc
Lesson 3
Describing Digit Manipulation
Overview
There are times when you might need to manipulate the digits ofthe telephone numbers that
come into and go out of your voice gateway. You may need to remove site codes for intersite
calls or add area codes and other digits for routing calls through the public switched telephone
network (PSTN). This lesson covers digit manipulation and digit manipulation tools.
Objectives
Upon completing this lesson, you will be able to describe how to implement digit manipulation
on a voice gateway. This ability includes being able to meet these objectives:
Explain how a gateway collects, processes, and consumes digits
Describe the components of digit manipulation and identify network points where digit
manipulation is applied
Describe digit stripping and its configuration options
Explain digit forwarding and its configuration options
Describe digit prefixing and its configuration options
Describe number expansion and how to configure il on a gateway
Describe CLID manipulation and its configuration options
Describe voice translation rules and profiles and explain how to implement them on a
gateway
Explain the capabilities ofthe dialplan-pattern command and contrast it to the voice
translation profiles
Describe how to test and monitor the functionality of digit manipulation on a gateway
Digit Collection and Consumption
This topic describes how gateways collect and process digits.
Digit Collection Met
Digit by Digit and En Block
1000
Dialed Digits
Destination patterns
1. .
10..
G>
&
&
Call Setup
Dialed Digits
|j^ j 1001 |
Potential matches
Potential matches
Potential matches
Potential matches
Find most explicit match
1001
Find the most explicit match
If an endpoint sends dialed digits otic by one. Cisco Unified Communications Manager Express
starts digit analysis immediately upon receiving the first digit.
By each additional digit that is received. Cisco Unified Communications Manager Express can
reduce die list of potential matches (that is. the call-routing table entries thai match the digits
that have been received so far). After a single entry, such as the directory number 1001 in the
example, is matched, the so-called current malch is used and the call is sent to the
corresponding dev ice.
Note Cisco Unified Communications Manager Express does not always receive dialed digits one
by one Skinny Client Control Protocol (SCCP) phones always send digit by digit. Session
Initiation Protocol (SIP) phones can use en bloc dialing to send the whole dialed string at
once, or can use Keypad Markup Language (KPML) to send digit by digit. If digits are
received en bloc, the whole received dial string is checked at once against the dial plan
Cisco Unified Communications Manager Express Addressing
Method
This topic describes how digit analysis is performed for different devices, according to their
addressing methods.
Manager Express Addressing Method
Signaling Protocol
Addressing Method
SCCP
IP phone
SIP
Gateway MGCP/SIP/H.323
Digit-by-digit
En bloc (Type-B phones only)
En bloc
KPML (Type-B phones only)
SIP dial rules
En bloc
OverlapsendinQ and receiving
(ISDN PRI only)
The table shows the addressing methods that Cisco Unified Communications Manager Express
supports for ditTerent devices.
In SIP en bloc dialing or KPML can be used. In en bloc dialing, the whole dialed string is sent
in asingle SIP INVITE message. KPML allows digits to be sent one by one. SIP dial rules are
processed inside the SIP phone. Therefore, aSIP phone can detect invalid numbers and play a
reorder tone, without sending any signaling messages to Cisco Unified Commun.cations
Manager Express. If dialed digits match an entry of aSIP dial rule, the dialed string is sent in a
single SIP INVITE message to Cisco Unified Communications Manager Express. If Cisco
Unified Communications Manager Express requires more digits, KPML can be used to send the
remaining digits one by one. from the SIP phone to Cisco Unified Communications Manager
Express.
ISDN PRIs can be configured for overlap sending and receiving, allowing digits to be sent or
received onebyoneoveran ISDN PRI.
Dial Plan Implementalion
User Input on SCCP Phones
mp:,:ScP^ hU CiSC Unined C^- ^**P processes user
4-56
User Input on SCCP Phones
Depending on the phone model and the way a number is
dialed. SCCP digit signaling can be en bloc or digit by digit.
Type-A phones (Cisco Unified IP Phone 7905, 7912, 7940
and 7960) support only digit-by-digit signaling.
Type-B Phones (all modern Cisco IP phones) support
enbloc anddigit-by-digit signaling.
Can becontrolled via the product-specific Enbloc
Dialing configuration parameter (enabled by default).
If the number isentered while the phone isonhook
and the Dial softkey is pressed to startthe call, enbloc
signaling takes place.
if the phone isplaced off hook first and then digits are
dialed, digit-by-digit signaling is used.
Whether anumber is signaled digit by digit or en bloc depends not only on the configured
signaling protocol but also on the phone model (Type Aor Type B) thai is used and on how the
phone number is dialed.
for Cisco SCCP IP phones, the following rules apply:
ivpe-A IP phones onlv support digit-by-digit signaling.
Type-B IP phones support digit-by-digit signaling as well as en bloc signaling.
Fn bloc dialing is used when acall is placed by the user entering the number while
the phone is on hook and then pressing the Dial softkey. Calls that are set up via
call-list entries orspeed dials also use en bloc signaling.
En bloc dialing, which is enabled bv default, can be disabled via the product-specific
Enbloc Dialing configuration parameter from the Phone Configuration page.
Digit-bv -digit dialing is used whenever Ihe number is dialed after the phone is put
off hook.
Note
The dialing behavior may vary based onthephone load version that is used.
SCCP Digit Collection
Thissubtopic explains digitcollection via SCCP.
SCCP Digit Collection
SCCP phones immediately report everyinput event (off-hook,
on-hook, each digit dialed, etc.) to Cisco Unified
Cisco Unified Communications Manager Express analyzes
phone input digit-by-digit against configured dial plan and
responds with feedback(dial tones, ringback, reordertone, etc.).
* Same as for analog endpoints on FXS ports.
sCCPmessage sent DiaiPlan .
y with each useraction (Digil Ana lysis) ^^
Off-hook, digil 1, digil0, dgitO, digit0
Dial tone on'olt, screen update, etc.
Dialing Actions.
1 000
An SCCP endpoint detects all userevents andrelays them totheCiscoUnified
Communications Manager Express individually. Auserwho goesoff-hook andthen dials 1000
would trigger five individual signaling events from thephone tothegateway. All theresulting
feedback that is provided totheuser, suchas screen messages, playing dial tone, secondary dial
tone, ringback, reorder, andsoon, arecommands that areissued bythe Cisco Unified
Communications Manager Express to thephone in response tothedial planconfiguration.
It is neither required norpossible toconfigure dial planinformation on CiscoUnified IPphones
running SCCP. All dial planfunctionality is contained in theCiscoUnified Communications
Manager Express system, including therecognition of dialing patterns as userinput is collected.
If the user dials a pattern that is deniedby CiscoUnifiedCommunications ManagerExpress, a
reorder tone is played to theuserwhen thatpattern becomes thebest match inCiscoUnified
Communications Manager Express digit analysis. Forinstance, if all callsto 92000 arcdenied,
a reorder tone would be sent to the user phone as soon as the user dials 92000.
SIP Digit Collection (Simple Phones)
This subtopic explains how simple (Type A) SIP phones report dialed digits.
Type A: Cisco Unified IP Phones 7905, 7912, 7940, and 7960
Phone accumulates all user input events until # or Dial
softkey is pressed (similar to cell phones)
Phone sends SIP INVITE message withcomplete dialed
digits (en bloc)
Cisco Unified CommunicationsManager Express analyzes
dialed digits against configured dial plan
Simple SIP Phone
SIP INVrTEmessage
/ senl when user presses
tne Dial key
"Call lor 1000"
Dial Plan
(Digit Analysis,
Call in progress, call connected, call denied, etc
Dialing Actions
100 0 Dial
With en bloc number reporting, the phone accumulates all user input events until the user
presses either the Uke\ or the Dial softkey. This function is similar to the dial button used on
mam mobile phones.
For example, a user making a call to extension 1000would have to press 1, 0, 0, and 0 followed
bv the Dial softke\ or the # key. The phonewouldthensend a SIPIN Villi, message to Cisco
UnifiedCommunications Manager Express to indicatethat a call to extension 1000is
requested. As the call reachesthe gateway, it is subjected to the dial plan, including all the class
of service and call-routing logic.
Implemenling Cisco Voice Communications and QoS (CVOICE) vS 0 2010 Cisco Systems, Inc
SIP Digit Collection (Enhanced Phones)
This subtopic explains how enhanced (Type B) SIP phones report dialed digits.
SIP Digit Collection (Enhanced Phones)
Type B: Cisco Unified IP Phones 7911, 7941, 7942, 7945,
7961, 7962, 7965, 7970, 7971, 7975, 7985.
Based on KPMLto report user key presses. Every key press
triggers a SIP NOTIFY message to Cisco Unified
Similar to SCCP and analog phones.
No Dial softkey to indicate the end of user input.
/
KPML events reported qî p|al
m SIP NOTIFYmessages (DigitAnalysis)
SIP Enhanced Phone .
Off-hook, BigIt1. digil 0. digit 0. digit0
Call in progress, call connected, call denied, etc.
Dialing Actions
1000
"fype B SIP phones offer functionalitythat is based on the KPMLto report user activities. Each
one ofthe user input events generates its own KPML-based message to Cisco Unified
Communications Manager Express. From the standpoint of relaying each user action
immediately to Cisco Unified Communications Manager Fxpress, this mode of operation is
similar to that of phones running SCCP.
Every user key press triggers a SIP NOTIFY message lo Cisco Unified Communications
Manager to report a KPML event corresponding to the key pressed by the user. This messaging
enables Cisco Unified Communications Manager Express digit analysis to recognize partial
patterns as they are composed by the user, and to provide appropriate feedback such as
immediate reorder tone if an invalid number is being dialed.
In contrast to fype A IP phones running SIP, Type B SIP phones have no Dial key to indicate
the end of user input. In the figure, a user dialing 1000 would be provided call progress
indication (either ringback tone or reorder tone) after dialing the last 0 and without having to
press the Dial key. This behavior is consistent with the user interface on phones running the
SCCP protocol.
) 2010 Cisco Systems, Inc. Dial Plan Implementation
Dial Peer Management
This subtopic explains the potential issues with dial peer matching when numbers are collected
digit-by-digit and compares digit-by-digit with en block collection.
Dial Peer fVlanagerr
Digit by Digit:
Example 1 Dialed dig ts 5550124 (one-by-or,e)
en1
ixample 2 Dialed digits 5550124 (one-by-one)
destination-pattern 555 0124
session target lpv4:10.18.0.2
destination-pattern 555 ....
session target ipv4:10.IB.0.1
Dial peer 1 will match first Onlythe Dial peer 2 will match first. The
collecteddigitsof 555 will be collecteddigitsof 5550124will be
forwarded forwarded
En bloc:
Examp.e 1 Dialed digils 5550124 (en block)
WiUi boh configurations above dial peer 2 will match Digits 5550124 will be foiwarded
The figure demonstrates the impact that overlapping destination patterns have on the call-
routing decision, fhe first two examples illustrate dial peer management with digit-by-digit
collection. In example 1. the destination pattern (555) in dial peer 1 is a subset ofthe
destination pattern (555....) in dial peer 2. With digit-by-digit number collection, the router
matches one digit at a time against available dial peers. This means that an exact match will
alwajs occur on dial peer 1. and dial peer 2 will never be matched.
In example 2. the length ofthe destination patterns in both dial peers is the same. Dial peer 2
has a more specific value than dial peer I. so it will be matched first. If the path to IP address
10.18.0.2 is una\ailable. dial peer 1 will be used.
Example 3 examines the dial peer management when the called number has been received en
block. Because the entire called number is available immediately, the second dial peer will
match in both configurations, because il oilers the most explicit match. The entire called
number (5550124) will be forwarded to the session target.
Digit Consumption
This subtopic describes how gateways consume digits when forwarding calls.
Digit Consumption
POTS dial peers:
- By default, the router consumes the left-justified digits that
explicitly match the destination pattern and forwards the
remaining digits.
- Use the nodigit-strip command to disable theautomatic
digit-stripping function.
VoIP dial peers: By default, the router forwards all digits
collected
Example 1; Dialeddigits 5550124
dial-poor voico 1 pots
pott 0/lil
Explicitly matched digts 555 are
consuned and 0124 is forwarded.
Example 2: Dialeddigits 5550124
dial-poet voice 1 pots
destination-pattern 555..
no digit-strip
port 0/1 !l __^
Digits5550124 are toiwarded.
Gateways process the called party number in two ditTerent ways, depending on whether the call
is forwarded over a POTS or a VoIP dial peer.
When agateway matches adial string to an outbound POTS dial peer, the router by default
strips off the left-justified digits that explicitly match the destination pattern. The remaining
digits are forwarded to the telephony interface, which connects devices such as aPBX or the
PSTN.
Digit stripping is the desired action in many situations. There is no need to forward digits out ol
a POTS dial peer ifit is pointing toan FXS port that connects atelephone orfax machine.
When digit stripping isturned offon this type ofport, the user will hear tones after answering
the call because any unconsumed and unmatched digits are passed through the voice path after
the call is answered.
When a PBX or thePSTN is connected through the POTS dial peer,digit stripping may not be
desired if these devices need additional digitstoroutethe call. Inthese situations, the
administrator must assess thenumber of digits that need to be forwarded for theremote device,
With aVoIP dial peer, all digits are passed across the network tothe tenninating router by
default. This behavior can be modified if required by the network.
Components of Digit Manipulation
Ihis topic pro\ ides an overview ofdigit manipulation mechanisms.
Digit Manipulation Mechanisms
Various mechanismsfor digit manipulations:
* Simple digit manipulation for dial peers:
digit-strip
forward-digits
prefix
did
* Global method to inflate or deflate numbers:
num-exp
Typically used for short dials and site codes
Voicetranslation rules and profiles:
Complexdigit manipulation using regular expressions
Translation rules used to manipulate calling party number,
called party number, or redirect number
Digit manipulation isthe task ofadding orsubtracting digits from the original dialed number to
accommodate user dialing habits orgateway needs. Digit manipulation incorporates adding,
subtracting, and changing telephone numbers. For example, you might need toadd the area
code to a call that will be routed out to the PSTNor remove a site code from an intersite eall
with the same compam . You can manipulate called numbers, calling numbers, redirected
numbers, and the number type. You can apply digit manipulation toincoming oroutgoing calls
or to all calls globally. You can manipulate digits before oralter matching adial peer.
fhe call agent performs digit manipulation inaMedia Cialeway Control Protocol (MGCP)
network: therefore digit manipulation may beperformed only onH.323 and SIP gateways.
Digit manipulation is an important aspect of any dial plan, andvarious toolsexistonCisco IOS
gateways to perform digit manipulation. Thesetools arc explained inthis lesson.
4-62 Implementing Cisco VoiceCommunicalions and QoS (CVOICE) v8 0
'2010 Cisco Systems, inc
Digit Manipulation Order
This subtopic describes theorder inwhich digit manipulation methods process signaled
numbers.
Digit Manipulation Order
POTS
POTS
1 Match outbound diaf peer
1] Diai-peer wice
trartalation profile
2) CLIO
3i Digit strip
4) PrefK cSgits
5) Forward digits
2. Outbound votca-pori
translation profile
The order of operation in digit manipulation follows the call through the gateway, as shown in
the figure. For inbound POTS calls, rules that are configured on the voice port are appliedfirst,
thenany global numberexpansion, followed by rulesthat are configured on the incoming dial
peer and thenthe outgoingdial peer. For inboundVoIPcalls, global voicetranslation profiles
are appliedfirst, thenanyglobal numberexpansion, followed by rules that are configured on
the incomingdial peer and then the outgoing dial peer.
Note The num-exp command is applied globally before any inbound dial-peer matching.
When possible, you should use a single method of accomplishing the required digit
manipulations. For example, do not use both the forward-digits and the prefix commands in a
dial-peer configuration.
It is possible to use all ofthe digit manipulation methods in a gateway. A single dial peer can
be configured with prefixes, voice translation rules, and clid commands. A call can be modified
by the voiceport, numberexpansion, inbounddial peer, and outbounddial-peerconfiguration
commands in a single gateway or multiple gateways. Understanding the order of operation in
digit manipulation is important, not only for configuration and tesl purposes but also for
troubleshooting.
Digit manipulationcan be very specific or global. If it is desired for all calls or required before
a dial peer is matched, then one ofthe global techniques is preferred. If digit manipulation is
needed for specific calls, it should be configured under the appropriate dial peers.
i 2010 Cisco Systems, Inc Dial Plan Implementation
Digit Stripping
This topic describes digit stripping.
IDPil
Dialed Number
^815551234
Dialefl Numtief:
911
dial-peer voice 911 pott
destination-pat tern
no digit-strip
PSTN
The digit-strip command is a dial-peer command that strips off the matched digits in a
destination pattern of a dial peer. The digit-strip command is supportedon POT'S dial peers
only. Digit stripping occurs aller the outbound dial peer is matched and before any digits are
sent out.
By default. POTS dial peers strip any outbound digits that explicitly match their destination
pattern, whereas VoIP dial peers transmit all digits in the called number. For example, given a
destination pattern of 5551.... the number thai is transmitted to the PSTN would contain the
last three digits. The first four digits. 5551. would be stripped because they explicitly match the
destination pattern.
In the example in the figure, users dial a 9 to reach an outside number. If the configured
destination pattern is 9T. then the 9 is matched and stripped from the called number that is sent
to the PSTN. On the other hand, you may have a dial peer for an emergency number such as
911 in the North American Numbering Plan (NANP) region. If the destination pattern is 911.
then \ou would not want the numbers stripped when they are explicitly matched. In this ease,
the no digit-strip command can be used to disable the automatic digit-stripping function. This
allows the router to match digits and pass them to the telephony interface. The figure shows an
example of such behavior.
Implementing Cisco Voice Communicalions and QoS (CVOICE| v8 0 Cg)2010 Cisco Systems, Inc
Digit Forwarding
This topic describes digit forwarding.
Digit Forwarding
Transmitted- NuffiBsr
1234
dial-psei voice 1000 pots
torward-diflitB 4
The forw.nl-dig.ta {num-digits | all | extra} command is adial-peer command, whidi
specifies how many matched digits should be forwarded. This command is available in POTS
dial peers only.
Digit forwarding allows more granular control over the number of transmitted digits Digit
fonvarding specifies the number of digits that must be forwarded to the telephony interface,
egardlcssoi whether they match explicitly or with wildcards. When aspecie number of ds
are configured for forwarding, the count is right-justified. For example, in the figure, th PO S
dial peer has adestination pattern that is configured to match all extensions mune 000 rang*
Bv default only the last three digits are forwarded to the PBX that is connected to the specified
voice port If the PBX needs all four digits lo route the call, you must use the command
forward-digits 4. or forward-digits all, so that the appropriate number of digits are forwarded.
Note
The forward-digits command is available in POTS dial peers only.
Digit Prefixing
fhis topic describes digit prefixing.
4-66
Digit Prefixing
dial-peer voice 2001
destination-patta rn 2
preference 1
prefix 3005553
port 0/1:23
PSTft-3
WAN-
The prefix command is adial-peer command that prefixes the specified digits to the number
fonvarded hx the dial peer. This command is applicable onlv to POTS dial peers When an
outgoing call ,s initiated to this dial peer, the prefix string value is sent to the telephonv
interlace first, before the telephone number associated with the dial peer.
Use the prefix command when the dialed digits leaving the router musl be changed from the
dialed digits that had originally matched the dial peer. For example, acall is dialed using a4-
digjt extension such as 2345. but the call needs to be routed to the PSTN, which requires 10-
d.git dialing. It the lour-digit extension mafchcs the last three digits ofthe actual PSTN
number you can use the prefix 3005552 command to prcpend the six additional digits and the
tl!f^,?ôTÂroftU8h,CXP,icit matCh- TheSC diehS arC needed for thc PSTN t0 *c
the call to ,00-555-2345. After the POTS dial peer is matched with the destination pattern of
2^45. the prefix command prepends the additional digits. The siring 100-555-^345 is then sent
out ot the \ oice port to the PSTN. " '
Implementing Cisco Voice Communicalions and QoS (CVOICE) v8 0
Simple Digit Manipulation Comparison
This subtopic pro\ idesa comparison of simpledigit manipulation mechanisms.
Simple Digit Manipulation Comparison
no digit-strip
digit-strip (default)
forward-digits 4
prefix 9
PSTN
Called patty number
913005551234
13005551234
1234
913005551234
The figure shows the operation of simple digit manipulation for POT'S dial peers. A user dials
9-1-300-555-1234, and the call is managed by the dial-peer voice 9 pots command on the
H.323 gateway. Depending on the commands, the called number will be modified differently:
If the no digit-strip command is used, the called number will be 9-1-300-555-1234. No
digits are modified.
If the digit-strip command is used, which is the default on all POTS dial peers, the
matched 9 will be stripped off, resulting in the called number 1-300-555-1234.
If the forward-digits 4 command is used, only the last four digits will be forwarded,
resulting in the called number 1234.
If the prefix 9 command is configured, the first digit (9) is stripped off and then prefixed
again, resulting in the called number 9-1-300-555-1234.
) 2010 Cisco Systems, Inc. Dial Plan Implementation
Number Expansion
Ihis topic describes global number expansion.
Number Expansion
Dialefl Number
5561234
num-exp 5551... 13005551...
no digit-strip
port 0/1:23
PSTN
300-555-1234
The num-exp command is a global command, which applies to all calls and performs a mutch-
and-replaceoperation to inflate or deflate called numbers. This command is typically used for
short dials and site codes. Number expansion occurs before matching a dial peer. Number
expansion provides an alternate method of adding digits lo outgoing calls. Whereas prefixing is
applied to a single dial peer, number expansion is applied globally to all calls, not just to calls
matching a single designated dial peer. Number expansionoccurs before the outbound dial peer
is matched, so \ ou must configure the outbound dial peer with the expanded number in the
The figure shows how the num-exp command expands the called number before outbound
dial-peer matching, 'fhe user dials 5551234. The gateway has the configuration num-exp
5551... B005551.... so the called number 5551234 is matched and modified to 1300-555-
1234. This called number matches dial peer 2000. which routes the call to the PSTN.
Note You can use the show num-exp command to view the configured number expansion table
Implementing Cisco Voice Communications and QoS (CVOICE! v8 i (2010 Cisco Systems, Inc.
Basic Digit Manipulation Example
This subtopic provides an example of basic digit manipulation.
Basic Digit Manipulation Example
1XXX 200555ixxx
num-exp 4_ 4005554..
dial-peer voice 4000 pots -**,
destination-pattern 4005554_
forward-digits all
prefix 1
port 0/1:21
dial-pear voice 3000 voip
session target ipv4 10.1.1.1
prefix 13005553
preference 1
port 0/1:23
no digit-atrip
port 0/1:23
forward-digits 4
The figure shows how to configure basic digit manipulation that meets the following network
requirements:
Users should be able to call remote sites using just the extensions for the specific sites:
3xxx and 4xxx respectively.
The PSTN should be used as abackup in case the WAN link isdown orcongested.
Users should be ableto contact 911 emergency services.
<2010 Cisco Syslems, Inc.
Caller ID Digits Manipulation
(his topic describes the methods of modifying calling line ID (CLID) informati
Caller JD Manipulation
routerlconfig-dial-peer)#
did {network-number number [second-number strip] |
network-provided | restrict | atrip [name] I substitute
name }
network-number option sets network number inCLID
second-number strip option removes second network
number from CLID information when setting primary number
network-provided option sets thescreening indicator touse
the number providedby the network
restrict option prevents the entire CLID (number and name)
from being presented
strip option removes the calling party numberfrom CLID
information
namekeyword removes additional calling party name
substitute nameoption copies thecalling number into the
display nameifthe calling name is empty
Cailing hue II) (CLID) is the collection ofinformation about the telephone number from which
a call originated. The CLID value might be the entire phone number, the area code, orthe area
code plus the local exchange. The did command has various keywords that manage (he
presentation, restriction, or stripping ofthe various CLID elements.
The did netnork-number command sets the presentation indicator to "y" and the screening
indicator to "network-pro^ ided." The second-number strip keyword strips the original callino
number from the H.225 source-address field, and is valid only ifanetwork number was
previously configured.
The did restrict command causes the calling-parly number to he present in the information
element, but the presentation indicator is set to "n" to prevent its presentation to the called
partv.
The didstrip command causes the calling-part} number to be null in the infonnation element
and the presentation indicator is set to "n" to prevent its presentation lo the called party. The
name keyword causes removal ofthecalling-party name from the CLID.
The did substitute name command copies the calling number into the display name ifthe
calling name is empty.
4-70 implementing Cisco Voice Communications and QoS (CVOICE) v8 0
Caller ID Manipulation Example
This subtopicprovides an exampleof caller IDmanipulation.
Caller ID Manipulation Example
Office
DID. 200-555-2xx*
192 168 1.1
On-Way VoIPCals
Home Office
DID: 300-555-1234
destination-pattern [1-8]...
did network-number 3005551234
The figure illustrates how to set the CLID number in the VoIP dial peer that is configured on
the home office gateway. When the home user calls an office number (any 4-digit number that
starts with 1-8), the call is sent over the Internet. The CLID number is set to the home office
PSTN number, which is presented as the calling number when the office phone rings. The
colleague in the office can call back the telecommuter over the PSTN. Such configuration is
helpful when the home office router has a dynamically assigned IP address. A dynamic IP
address cannot be referenced by Ihe static VoIP dial peers configured on the office router.
>2010 Cisco Systems, Inc. Dial Plan Implementation 4-71
Digit Manipulation Using Voice Translation Rules
and Profiles
4-72
fhis topic describes \oice translation rules and profiles.
Voice Translation
Voice translation rules define up to 15 sub-rules to
manipulate digits. TONs, and numbering plans.
Voice translation profiles reference up to three rules:
Called: Translation rule for the called number
- Calling: Translation rule for the calling number
Redirect-called: Translation rule for the redirect number
Voice translation profiles can be referenced by.
VoIP dial peers, voice ports, any inbound VoIP call, a
specific range of source IP addresses in VoIP calls, trunk
groups, NFAS controllers, or SRST
Voice translation rules and profiles are the most powerful Cisco IOS Software tools you can
use to perform digit manipulation. Using regular expressions, a numbering plan, and type-of-
number (TON) matching, you can make nearly any possible modification to digits. Such
modifications can include manipulating the calling party number, manipulating the called party
number, or changing the redirect number digits for a voice eall. The only drawback is the
complex syntax.
The voice translation rules are associated with a voice translation profile, which can reference
up to three \oiee translation rules as follows:
A voice translation rule that is used to manipulate the called number
A voice translation rule that is used to manipulate the calling number
A voice translation rule that is used to manipulate the redirected called number
The resulting voice translation profile can be attached lo these elements, either for incoming or
for outgoing calls, as follows:
VoIP dial peers
Voice ports
Any inbound VoIP call
A specific range of source IP addresses in VoIP calls
A trunk group
A Tl/I. I controller that is used lor Non-Facility Associated Signaling (NFAS) trunks
Survivable Remote Site lelephony (SRS'I )
Implementing Cisco Voice Communications and QoS (CVOICE! "8 0 2010 Cisco Systems, Inc.
Voice Translation Rules and Profiles Hierarchy
This subtopic explains the hierarchy of avoice translation profile and rule configuration.
Voice Translation Rules and Profiles
Hierarchy
1 7 5 4 6
*
- n
M 1? IS *4 15
1 a 1 *
- b D 1C
li
1 '2 -3
1-
a s 4 E
- 8 P 1"
1-
z 12 '4 *
The figure illustrates the concept ofvoice translation profiles and rules. Each voice translation
rule can have up to 15 individual subrules. Amaximum of128 translation rules are supported.
The \ oice translation rule isthen referenced by a voice translation profile for called, calling,
and redirected called number.
Note lhat the same voice translation rule can bereferenced bymultiple voice translation
profiles.
Note Although you can have up to 15 subrules within a voice translation rule, the first matching
rule will beapplied and no further subrules will beconsidered. __
Voice Translation Rule Regular Expressions
This subtopic describes the regular expressions that arc used by voice translation rules.
Voice Translation Rule Re<
* Matchthe expressionst the start of a line,
$ Malchihe expression at Ihe end of the line
/ Delimiter marking start and end ofboth matching and replacement strings.
\ Escape Ihe special meaningofthe nextcharacter.
Indicates a range when used withinbrackets.
[list] Match a single character in a list.
['Hist] Donotmatch a single characterspecified inIhelist
M3tch any single character
Repeattheprevious regular expression zeroor morelimes.
+ Repeat the previous regular expression one or more times
, Repeat theprevious regular expression zerooronelime (useCtfl-V inorder
to enter in Cisco !OS Software)
() Groups regular expressions UseV1-9 torefer lomatched groups.
& Match thesubstring (matched string) You may alsouse\0.
Voice translation rules use regular expressions for match-and-replace operations, "fhe table
describes themost important regular expressions.
Regular Expressions for Voice Translation Rules
Voice Translation
Rule Character
[list]
Alist]
Description
Match the expression at the start of a line
Match the expression at the end of the line.
Delimiter that marks the start and end of both the matching and replacement
strings
Escape the special meaning of the next character.
Indicates a range when not in the first/last position. Used with '[' ,and ']
Match a single character in a list.
Donot matcha single character that is specifiedinthe list.
Match any single character
Repeat the previous regular expression zero or more times
Repeat the previous regular expression one or more times.
Repeat the previous regular expression zero times or one time
Groups regular expressions.
4-74 Implementing CiscoVoice Communications and QoS (CVOICE) v8.i
>2010Cisco Systems, Inc.
Voice Translation Rule Operations
Thissubtopic explains voice translation ruleoperations.
Voice Translation Rule Operations
PSTN-Out
The rule rule 1 /rt821 S/ /3005551 / says
match 82 I.
Site A, DID: 200-555-3xxx
1xxx
User dials 821001 to
reach a Site B user but
call goes through PSTN
and change to 3005551
SiteB
Intersite Prefix: 8
Site Code: 21
DID; 300-555-1xxx
PSTN-ln
The rule rule 1 /A2005553 S/ /1 / ^VS
match 2005553 and change to 1..
Use the voice translation-rule command to create the definition of a translation rule. When the
router evaluatesa translation rule, it is reallyonly performing a "matchthis and changeto this"
operation on the regular expression.
Consider the following examples:
This rule will be used to change die outgoing called number to a 10-digit number for
routingacrossthe PSTN. The rule will be appliedoutgoingon an interface, port, or dial
peer,
Router(config)# voice translation-rule PSTN-out
Router(config)# rule 1 /A821...$/ /3005551.../
This rule will be used to change the incoming calling number to a four-digit number after
routingacrossthe PSTN. Therule will be appliedincomingon an interface, port, or dial
peer.
Router(config)# voice translation-rule PSTN-in
Router(config)U rule 1 /"2005553...$/ /l.../
This table shows ihe match-and-replace operations for each rule:
Rule Match This Change To
/"821 S//3005551 ../ 821... 3005551...
/A200553...$//!.../ /2005553.../ /!../
>2010 Cisco Systems. Inc Dial Plan Implementation
Prepending Digits
fhis subtopic explains howto use voice translation rules for prepending digits.
ing Digits
\ = escape character
User sees
93005551234 m
thecal! lisl and
can call hack
Calling number:
93005551234 Gateway prepends 9 to
incoming calling number
for calling back
-PSTW-
Calling number.
3005551234
3005551234
Voice translation rules provide a handy way to prepend a 9 to all outgoing calls. It would not be
feasible to use indi\ idual translation rules for each number because ofthe number of rules
needed. The following is an example of voice translation rules:
rule 1 /30C5550100/ /93005550100/
rule 2 -'3005550101/ /93005550101/
rule 3 ''3005550102/ /93005550102/
Using a \ariable simplifies the configuration effort. Translation rule expressions can be divided
into sections b> using an "escape"" character to create these variables. The regular expression
escape character is "V.
For example. \ou may use the following translation rule lo prepend a 9 lo outgoing calls for
routing through the PS'I'N:
rule 1 /$"[2-9] $/ /9\l/
This rule would prepend a 9 to whatever was matched in the first set of parenthesis (\l). In
other words, it would replace \! with A[2-9j ) and add a 9 to the beginning.
Voice Translation Rule Search-and-Replace Examples
This subtopic provides common usage examples of voice translation rules.
Voice Translation Rule Search-and-
Replace Examples
1*911/
/*2001//3001/
/1231...//4D0a
/"2...//801SJ
/*2...//801\Qf
/.V/91&/type national
national
& ix '0 rnii he used to
replace Iheoriq'na! digits
in out
mj _ 913005552001 type
3005552001 type national nMQMi
The table lists sonic search-and-replace operations that use voice translation rules.
Examples of Voice Translation Rules
Rule
Input String
Output String
/A9/ //
913005550123
13005550123
/A2001//3001/ 2001
3001
/*[23]...//4000/ 2025 or 3051 4000
/A2...//801&/ 2001
8012001
/A2...//80t\0/
2001
8012001
A(9\)$[A10].*$/A114Q8\2/ 95551234
914085551234
/,*/ /91&/ type national national
3005552001 type national 913005552001 type national
Voice Translation Rule Search-am
Replace Examples (Cont.)
Translation Rule: A(9\)$[A01 ].*$/ A1130012/
Search
Replace
1300
Input
Output
1300
This example shows acomplex search-and-replacc operation in which this rule is configured:
rule 1 /$9$$["01].*$/ /\11300\2/
Theexample would begood for prepending a long distance T' andanareacodetoa dialed
number that is exiling die network viathePSTN andaccessing a long-distance subscriber, 'fhe
user would be dialing a 9 plus seven digits to access outside numbers.
If the input string95550134 is used, the operation proceedsas follows:
Step 1 The 9 will be reinserted using the \l.
Step 2 This entry is followed by the digits 1300.
Step3 fhis entry is thenfollowed by 5550134. whichis referenced by the \2.
Step 4 Theresulting string would be 913005550134.
Note
Tne first set of parentheses is referenced as \1 and the second set as \2
Assigning Rules to Profiles
This subtopic explains how toassign rules to profiles.
Assigning Rules to Profiles
Attribute Description
called Definesthe translation profile rule forthe
called number
calling Defines the translation profile rule for the
calling number
redirect-caled Definesthe translation profile rule for the
redirect-called number
The voice translation profiles collect asetof rules that, taken together, translate the called,
calling, and redirected numbers in specific ways. Up to 1000 profiles can be defined. Each
profile must have a unique name. After atranslation profile isdefined, it can be attached to
these elements:
Trunk Group: Two different translation profiles can bedefined ina trunk group inorder
to perform number translation for incoming and outgoing POTS calls.
Source IPGroup: Atranslation profile can beattached toa source IPgroup inorder to
perform number translation for ineoming VoIP calls.
Dial Peer: Twodifferent translation profilescanbe definedin a dial peer in order to
perform number translation forincoming and outgoing calls.
Voice Port: Two different translation profiles can bedefined ina dial voice port inorder to
perform number translation for incoming and outgoing calls. If a voice portis also a trunk
group member, then the incoming translation profile ofa voice port overrides the
translation profile of a trunk group.
NFASInterface: Thetranslation profilecanbe attachedto an NFAS interface.
Translation Profile Processing Order
This subtopic discusses the processing order of multiple translation profiles that are attached to
various elements.
Translation Profile Processing Order
JPOTS J POTS 1POTS J VoIP
Voice Port Trunk Group NFAS Interface Source Group
4
Inbound Dial Peer
\
Outbound Dial Peer
Ipots j pots 4pots I
Voice Port Trunk Group NFAS Interface I
voip
|P0TS JPOTS JPOTS
When a POTS call arri\es ongateway, the gateway tirst checks the incoming voice translation
profile ofthegiven voice port. Ifthe incoming call arrives through a port without anincoming
translation profile but belongs toa trunk group with such a profile, thetrunk group profile is
used. If the POTS call arri\es overan NFAS interface, its incoming profile is applied. The
general ruleisthat themore specific setting overwrites themore generic configuration. If the
call arrives as VoIP, the incoming Source (irouptranslation profile is applied tothecall if the
source address ofthe VoIPcall is matchedbv the SourceGroupdefinition.
Next, the incoming translation profile that is applied tothe inbound dial peeris processed,
followed bv the outgoingtranslation profile inthe outbound dial peer. If the call is forwarded as
a POTS call, it is still processed by the outgoing translation profilelhal is configured on the
givenvoiceport, trunk group, or NFAS interface. The voiceport settingoverrides the trunk
group configuration, if both are available.
Incoming PSTNCall Example 1
This subtopic provides an example of called number translation for ineoming PSTN calls
Incoming PSTN Call Example 1
Translation Profile Applied to Inbound Dial Peer
It
voice translation-rule 1
rule 1 /-2005552/ /I/
I
voice translation-profile pstn-io
translate called 1
I
translation-profile incoming pstn-in
incasing called-number .
DID.2D05552XXX
VV
PSTN
Phone rugs
Profile moaties called number
to 1001
User dials
12005552001
The example inthe figure illustrates how voice translation profiles are used totranslate the
called number.
The incoming call is processed inthe following order:
Step 1 A PSTNuser dials 1-200-555-2001.
Step 2 The call isrouted toCisco Unified Communications Manager Express, which has a
DID range of 42005552XXX. Thevoice translation profile modifies the called
number to 1001, which matches the dial peer ofthe registered phone.
Step 3 The Cisco Unified IP phone rings.
Incoming PSTN Call Example 2
This subtopic provides an example oftranslating both the calling and called number for
incoming PSfN calls.
Incoming PSTN Call Example
Translation Profile Applied to Voice Port
voice trail elation-rule 1
rule 1 /* 2005552/ 12!
voice ttanElation-rule 2
rule i r . / /9s/ type sub scriber nub riber
rula 2 / .'/ /SIS/ type na tional r ati ll
rule i i .*/ /9011s/ type nterna ional international
voice tranelation-profile p tn-in
trail slate called 1
ttan late calling 2
voice port 0/0/0:15
tran lati on-profile incoming pstn in
Inthisexample, the\oice translation profile is required toperform the following
manipulations:
The incoming called number 2005552XXX should be modified to 2XXX,
fhe incoming calling number should be prefixed with theappropriate PSTN access code
and identifier;
local calls: Prefix 9
National calls: Prefix 91
International calls: Prefix 9011
An inbound PSTN call is processed as follows:
Step 1 A PSTN user dials l-200-555-2001 from 300-555-0123.
Step 2 The gatewav accepts the call and modifies the calling and called numbers. The rule
/A2005552/ IIImodifies thecalled number to 2001. and therule/W/9I&/ type
national national modifies the calling number to 9-1-300-555-0123.
Step 3 The internal phone rings.
2010 Cisco Syslems. Inc
Incoming PSTN Call Blocking Example
This subtopic provides an example of how to selectively block incoming PSTN calls.
incoming PSTN Call Blocking Example
Block calls from PSTN area code 399-555
rule 1 reject /'3S9555/
voice translation profil* bloc*
translate calling 1
call-block translation-profile incoming block
call-block disconnect-cause incoming invalid_number
Calls can beblocked when arriving onthe gateway. Outgoing calls cannot beblocked. From
the perspective ofthe gateway, the incoming direction can be either ofthe following:
Incoming from a voice port
Incoming over an inbound VoIPcall froma peer gateway
In this call-blocking example, the gateway blocks any incoming call that successfully matches
inbound dial peer 1and has a calling number that starts with 399555. Acomponent ofthecall-
block command isthe ability toreturn adisconnect cause. These values include call-reject,
invalid-number, unassigned-number, and user-busy. In this example, when the dial peer 1
matches acall and the calling number starts with 399555, the gateway will reject the call and
return a disconnect cause of "invalid number' to the source ofthe call.
Digit Manipulation Using dialplan-pattern
Command
This topic describes digit manipulation using the dialplan-pattern command.
Digit Manipulation Using dialplan-
pattern Command
routerlconfig-telephony)#
dialplan-pattern tag pattern extension-length extension-length
[extension-pattern extension-pattern | no-reg] [demote]
* Expands extension numbers into fullyqualified E 164 numbers
Creates another dial peer for everySCCP and SIP directory
number
pattern. PSTN area code, prefix, and the first one or two digits of
the extension number, plus dots (.)for remainingextension digits
no-reg option prevents the E.164 numbers from registering with
gatekeeper
demote keyword enables "+"E.164 inbound dialing
Availablefor SCCP (telephony-service) and SIP (voice register
global) Cisco Unified Communications Manager Express endpoints
* Does not apply for analog endpoints attached to FXS ports
Alternative voice translation profiles
The dialplan-pattern command, available inCisco Unified Communications Manager F.xpress
mode fbr SCCP (telephony-service) and SIP (voice register global) endpoints. creates a pattern
forexpanding internal extensions into fully qualified E.164 numbers.
The command maps internal extensions toDID numbers. This mapping isdone by dynamically
creating a new dial peer that has the DID number ol'a phone asthedestination pattern. Ihis dial
peeris alsoused for outbound callsto present thecorrect calling number andcanbe used to
register the full DIDnumber with a gatekeeper.
If multiple dial-plan patterns aredefined, thesystem matches extension numbers against the
patterns insequential order, starting with the lowest numbered dial-plan pattern tagfirst. Once
a pattern matches an extension number, the patternis usedto generatean expanded number. If
additional patterns subsequently match the extension number, theyare not used.
Thenumber expansion does not cover extensions of KXS ports andvoice-mail pilots.
fhe demote option enables the internal endpoints tobereached from PS'I'N using L. 164
numbers with theprefix. Thedialplan-pattern command is then used in theopposite wav.
because the endpoints are configured using L.164numbers, and as such can be reached from
the PSTN. The internal users, however, candial each other by usingshorterextensions defined
by the extension-length kevword.
As an alternative to the dialplan-pattern command, voice translation profiles can be
configured to mapthe internal extensions to DIDnumbers. The voicetranslation profiles,
however, are more complex and do not register theexpanded numbers onthegatekeeper.
4-84 Implemenling Cisco Voice Communications and QoS (CVOICE) vS.O
Digit Manipulation Using dialplan-pattern Command Example 1
This subtopic presents an example ofhow to use the dialplan-pattern command.
pattern Command Example 1
Leading Extension Digits Match DID Range
voice register global
aialplan-pattera 1 2005552... extanaion-langth 4
I
telephony-service
dialplan-pattsro 1 2005552... extension-length 4
The figure shows how to use the dialplan-pattern command when the leading extension digits
match the DID range,
An incoming PSTN call is processed in the following order:
Step1 A PSTN user dials 1-200-555-2001.
Step 2 The call isrouted toCisco Unified Communications Manager Express, which has a
DIDrange of 2005552XXX.
The gateway matches the outbound dial peer for the called number 200-555-2001
that is automatically created by the dialplan-pattern command for Ihe registered
phonewithextension2001.
Step 3 The registered phone with extension 2001 rings.
Note Analog phones thatare connected to FXS ports are notservedbythiscommand.
Ihis subtopic provides an example for using the dialplan-pattern command with the
extension-pattern option.
Leading Extension Digits Do Not Match DID Range
voice register global
dialplan- pattern 1 2005552... extension- length 4 extension-pattern 1
I
tele phony-service
dialplan-pattern 1 2005552... extension-length 4 extension-pattern 1.,
The figure illuslrates how to use the dialplan-pattern command when the leading extension
digits do not match the DIDrange.
The difference from the prev ious example is inhow ihegateway infiates the number. Instead of
just prepending the DID prefix, the internal range IXXX is mapped tothe DID range
2005552XXX,
4-86 Implemenling Cisco VoiceCommunicalions and QoS (CVOICE] v8.0
mm
This subtopic provides an example for using the dialplan-pattern command to support
inbound E. 164 "+" dialing.
Dialing Using + Prefix
i/oica register global
dialplan-pattern 1 42005552... extension-length 4 denote
telepbony-aervice
dialplan-pattern 1 2D0S552... extanaion-length 4 demote
o
Phone mgs
+2005552001 DO *2005552XXX
Internal endpoints can cal
each other using internal
extensions, such as 2001
Vv
Dialplan-pattem does nol
modify the called number
PSTN
y\_
User dials
2005552001
This figure illustrates how the demote option enables the PSTN toreach the Cisco Unified
Communications Manager Express endpoints using E.164 withthe+ prefix.
The dialplan-pattern command can beused intheopposite way byadding thekeyword
demote to the end ofthe command. In this case it demotes IP phone directory numbers that are
specified inE.164 fonnat with a+prefix toshorter extensions, which are tobeused internally.
External callers placecalls to thephones using E.164 format with a + prefix. If thecallsare not
natively received inthis format from thePSTN you have totransform thecalled number using
other methods, suchas translation profiles. Internal users, however, can dial eachother by
using shorter extensions, which areset upbythedialplan-pattern command.
Inthis example, thephones areconfigured with directory numbers +200555.... and have to be
called like that from the outside. Internal users, however, can call each other by using the last
four digits,
>2010 Cisco Syslems, Inc.
Verifying Digit Manipulation
This topic describes hov\ tovcrifv digitmanipulation.
Verifying Digit Manipulation Overview
Command
showdial plan number
show voice translation-profile
show voice translation-rule
test voice translation-rule
Description"
Displaysthe matching outgoingdial peer for
a telephone number
Displays all or selected voice translation
profiles
Displays all or selected voice translation
rules
Tests Ihe operation of a voice translation rule
for a specific telephone number
Cisco Unified Communications Manager Express allows verification and testing of digit
manipulation using the show dialplan number, show voice translation-rule, show voice
translation-pro file, and test voice translation-rule commands.
Implementing Cisco Voice Communicalions and QoS (CVOICE] v8 0 2010 Cisco Systems, Inc
II
Verifying Dial Plan
This subtopic explains how toverify the dial plan.
Verifying Dial Plan
router# show dialplan number 913005551234
Macro Exp.: S13005551234
VoiceBncapPeer91
peer type - voice.
information type = voice.
description = ~',
tag = 91, destinati
answer-address = ~' , preference*!).
CLID Restriction - Nona
CLID Network Dumber = '5551234'
CLID Second Number sent
CLID Override RDNIS . disabled.
The show dialplan number command displays which outgoing dial peer isreached when a
particular telephone number is dialed. This command is useful for testing whether the dial plan
configuration is valid and working as expected. Itincludes various additional parameters,
including the CLID options that aresetfora particular dial peer.
Verifying Translation Rules and Profiles
This subtopic describes how toverify translation rules and proliles.
router* show voice translation r le 1
Tiansllticn-ruls tag: 1
Rule ::
Match pattern: *55S\I.
\)
Replace pattern; 444V1
Ma ten type: none
Replace type: none
Match plan: none
Replace plan: none
Rule 2
Match pattern: 777
Replace pattern: 8S8
Match type: national Replace type: unknown
Match plan: any
Replace plan: ifidn
router* show voice translation- profJle
Translation Profile: swap prefi X
Rule (ot Calling number
Rule for Called number: 1
Theshow voicetranslation-rule command displavs oneor more translation rules. Il canbe
used for viewing a particular translation rule or all translation rules that aredisplayed in
ascending or descending order. The output also lists the numbered subrules.
The show voice translation-profile command displays one or more translation profiles. It can
be used for viewing a particular translation profile orall translation profiles that are displaved
in ascending or descending order.
4-90 Implemenling Cisco Voice Communications and QoS (CVOICE] vB.O
'2010 Cisco Systems, Inc
Testing Translation Rules
This subtopic explains how totest translation rules.
Testing Translation Rules
router* test voice translation-rule 5 2015550101
Matched with rule 5
Original numberi2015550101 Translated number:1025550101
Original number type: none Translated number type: none
Original number plan: none Translated number plan; none
router* test voice translation-rule 5 2125550101
2125550101 Didn't match with any rules
The test voice translation-rule command is usedto test the functionality of a translation rule.
'fhe complete command syntax is as follows:
test voicetranslation-rule number input-test-string [typematch-type [plan match-type] ]
The number type and calling plan are optional parameters that are defined ina translation rule.
Ifeither parameter isdefined, the call must match the match pattern and the type orplan value
in order to be selected for translation.
Validvalues for the type match-type argument are as follows:
abbreviated: Abbreviated representation ofthe complete number
any: Any type of called number
international: Number called to reach a subscriber in another country
national: Numberto reacha subscriberin the samecountry but outsidethe local network
network: Administrative or service number specific to the serving network
reserved: Reserved for extension
subscriber: Number called to reach a subscriber in the same local network
unknown: Number of a type lhat is unknown to the network
Valid values for the plan match-type argument are as follows:
any: Any type of called number
data: Number called for data calls
ermes: European Radio Message standard numbering plan
isdn: Called number for an ISDN network
national: Number called toreach a subscriber inthe same country buloutside the local
network
private: Number called for a private network
reserved: Reserved for extension
telex: Numbering plan lor telex equipment
unknown: Number of a type thai is unknown to the network
4-92 Implementing Cisco Voice Communications and QoS (CVOICE) v8 0 2D1D Cisco Systems. Inc
Summary
Summary
Thegateway typically collects digits ona digit-by-digit basis.
En bloc collection is characteristic for SIP INVITE messages
from simple SIPphonesand call setupmessages received
over digital PSTN trunks.
Digit manipulation canbe applied globally, toVoIP calls
originated from specific IP addresses, todial peers(inbound
and outbound), and voice ports(inbound and outbound).
Digit stripping in called numbers is controlled by thedigit-
strip and forward-digits command.
1The forward-digit command defines how many digits ofthe
called number are forwarded and has different defaults for
POTS and VoIP dial peers.
Summary (Cont.)
The prefix commandis used to prepend a prefix to the called
number.
Number expansion convertsa particular set of numbers into
a defined destination pattern.
The did command is used to modifycalling line ID
information.
Voice translation rules and profiles are used to manipulate
the calling, called, and redirect numbers.
The dialplan pattern commandadds fully qualified E.164
numbers for registered Cisco Unified Communications
Manager Express endpoints.
Verification ofdigit manipulation may involve the viewing of
the dial plan and testing of translation rules.
4-94 Implementing Cisco Voice Communicationsand QoS (CVOICE) v8.0 2010 Cisco Systems, Inc.
mm
Lesson 41
Configuring Path Selection
Overview
Path selection isoneofthe most important aspects of a well-designed VoIP system. High
availability is desirable so that there is usually more than one path for acall to take to ils final
destination. Multiple paths provide several benefits, including redundancy in case ofa link
failure or insufficient resources on that linkand a reductionin toll costs of a call. This lesson
introduces the path selection strategies and methods toimplement them.
Objectives
Upon completing this lesson, you will be able todescribe how agateway can be configured to
perform path selection. This ability includes being able tomeet these objectives:
Describe how the voice gateways select thecorrect path when routing voice calls
Explain how agateway matches dial peers todetermine path selection
Describe die variouspathselection strategies
Describe the characteristicsof site-code dialing and toll bypass
Describe howto configure site-codedialingand toll bypassin a gateway
Explain theprinciple andcharacteristics of TF.110
Describe how to configure TEHO
This topic provides an overview ofcall routing and path selection.
Relies on dial peers
Route to;
Digital or analog voice circuits
- VoIP peers (H.323 or SIP)
Registered Cisco Unified IP phones
* One dial peer associated with each call leg
Thecall-routing logicon Cisco IOS gateways relieson the dial-peerconstruct. Each cal!
passing through the Cisco IOS router isconsidered tohave two call legs, one entering the router
and one exiting the router, fhe call legentering therouter istheincoming call leg. while the
call legexiting the router is the outgoing call leg.
fhere are two main tvpes of call legs:
Traditional telephony calllegs: Connect tothepublic switched telephone network
(PSTN). analog phones, or PBXs
IP call legs: Connect the routerto other gateways, gatekeepers, or Cisco Unified
Communications Manager F.xpress systems
Dial peers are of two main types:
Plain oldtelephoneserv ice(POTS) dial peers: Associated with traditional telephony call
legs
VoIP dial peers: Associated v\ith IP call legs
Cisco Unified Communications Manager Express implements theendpoints differently
depending on the signaling protocol, as follows:
CiscoUnified Communications ManagerF-xpress createsPOTS dial peers for Skinny
Client Control Protocol (SCCP) endpoints. Therefore, (heir call legs are considered POTS,
although SCCP is a VoIP protocol.
Cisco UnifiedCommunications Manager Express creates VoIPdial peers for Session
Initiation Protocol (SIP) endpoints. Communicationbelween SIP phones and other VoIP
dev ices musl be explicitly permitted by the allow-conneetions sip to h323 and allow-
connections sip to sip commands in voice service voip mode.
Dial-Peer Matching
This topic describes how dial-peer matching isrelated topath selection.
Matching Dial Peers
Incoming Outgoing
Call Leg Call Leg
\ /
Incoming Outgoing
Dial Peers Dial Peers
H.323/SIP Gateway
Gateways must match Ihe correct inbound and outbound dial peers tosuccessfully complete a
call. Eor all calls going through thegateway, theCisco IOS gateway associates onedial peerto
each call leg. The figure shows examples of different types of calls going through a Cisco IOS
gatewav'.
Dial Peer Matching Refresher
This subtopic prov idesa refresher of dial-peer matching.
Dial-Peer Matching Refresher
Inbound dial-peer matching:
Called numberwith incoming called-number
.':. Calling numberwith answer-address
Calling number with destination-pattern
'-. For POTS: voice-port matches with dial-peer port
": Still no match: default dial peer 0 is used
Outbound dial-peer matching:
i Gateway tries to match the called number with
destination-pattern
7 By default, if multiple matches are found, the best
(numerically lowest) preference wins
s By default, if equal preferences are found, a random dial
peer is chosen.
The figure provides a rev iew of inbound andoutbound dial-peer matching.
Inbound Dial-Peer Matching
Inbound dial-peermatching for digital POTS and VoIPdial peers is prioritized as follows:
1. If thecalled number matches theincomingcalled-numberconfiguration on a dial peer,
thisdial peerwill beselected as theinbound dial peer. Nofurther matching is perfomied.
2. If no dial peer has been found, the calling number is checked. If the answer-address
configuration of a dial peer is matched, this dial peer will be seiecledand no further
matching is performed.
3. If the callingnumbermatches vsith the destination-pattern configuration of a dial peer,
this dial peer uill be selected and no further matching is performed.
4. If none ofthe above were successful and thecall is inbound on a POTS port, a dial peer
v\ith a matching voice-port configuration is searched.
5. If a match is still not found, the default dial peer 0 is used.
The routerneeds lo matchonlyone of theseconditions. It is nol necessary for all theattributes
to be configured in the dial peer or that every attribute match Ihc call setup infonnation. The
router stops searching as soon as one dial peer is matched, and the call is routed according lo
the configured dial-peerattributes. Evenif there are other dial peersthat wouldmatch, onlythe
first match is used.
Note Atypical misconception about inbound dial-peer matchingis that the session-target of a
dial peer is used This is not true, instead, use the incoming called-number or answer-
address command to ensure that the correct inbound dial peer is selected.
2010CISCO Systems. Inc.
Inbound dial-peer matching for dial peers pointing to analog voice ports is prioritized as
follows:
1. If thestation-id number has been set onthe portandif the calling number matches withthe
destination-pattern configuration ofadial peer, this dial peer will be selected and no
further matching is performed.
2. If the abovewas not successful and the call is inbound on a POTS port, a dial peer witha
matching voice-port configuration is searched.
Outbound Dial-Peer Matching
Outbound dial-peermatching is prioritized by default as follows:
1. The gateway searches through all dial peers and tries tomatch the called number with the
destination-pattern configuration. Thedial peerwith themost specific match isselected.
2. By default, if multiple equal matches arefound, thedial peerwith thelowest preference
configuration wins.
3. Bydefault, if equal preferences are found, a random dial peeris selected.
12010 Cisco Systems, Inc Dial Plan Implementation 4-99
Matching Dial-Peer Commands Refresher
This subtopic prov ides arefresher ofthe dial-peer commands relevant for dial-peer matching.
ching Dial Peer Comi
router(con fig-dial-peer I#
incoming called-number 1+]sainglT]
Specifies the incoming called number that will be used during
inbound dial-peer matching
router(contig-dial-peer)#
answer-address [+] stringlT]
Specifies the incoming calling number that will be used
during inbound dial-peer matching
router(con fig-dial-peer|#
destination-pattern [ +] st.ri/ig[T]
Defines the destination pattern of a dial peer that will be used
during inbound and outbound dial-peer matching
"Ihe figure shows commands that are used to configure calling andcalled number matching on
dial peers.
Calling and Called Number Matching on Dial Peers
Command Description
destination-
pattern
[ +]stringlT]
Use this command in dial-peer configuration mode to specify
either the prefix or the full E.164 telephone number to be used
for a dial peer. To disable the configured prefix or telephone
number, use the no form of this command.
incoming
called-number
[ +]scring[T]
a digit string that can be matched by an incoming call to
associate the call with a dial peer. To reset to the default, use
the no form of this command.
answer-
address
[+]stringlT]
the full E.164 telephone number to be used to identify the dial
peer of an incoming call. To disable the configured telephone
number, use the no form of this command
4-100 Implementing Cisco Voice Communications and QoS (CVOICE] v8.0 2010 Cisco Systems. Inc
Matching Dial Peer Commands (Cont)
direct-inward-dial
Enables DID on inbound POTS dial peers
Used to prevent two-stage dialing
preference [0-9]
Specifies the preference of a dial peer; default =0
router(config)#
no dial-peer outbound status-check pots
Disablesstatus checkingof outbound POTSdial peers duringcall setup
Includes POTS dial peers in call routing, even if status is down
Useful for some ISDN BRI links that deactivate Layer 1 during inactivity
The figure shows commands that areused toconfigure direct inward dialing (DID), dial-peer
preferences, and outbound status checks.
direct-inward-dial and dial-peer matching Commands
Command Description
direct-
inward-dial
Use this command to enable the DID call treatment for an
incomingcalled number. When this feature is enabled, the
incoming call is treated as ifthe digits were received from the
DIDtrunk. The called number is used to select the outgoing
dial peer. Nodial tone is presented to the caller.
preference
[0-9]
Use this command in dial-peer configuration mode to indicate
the preferred order of a dial peer within a hunt group. To
remove the preference, use the no formof this command. The
default is 0 and is not displayed in a configuration.
no dial-peer
outbound
status-check
pots
Use this command in privileged EXEC mode to disable the
checking of the status of the outbound POTS dial peers during
call setup and to allow for that call any dial peers whose status
is down.
This may be required on some ISDN links where the central
office (CO) ISDN switch activates the ISDN layer only if activity
is detected on the link.
Outbound Dial-Peer Matching Order
This subtopic explains how todefine the matching order for outbound dial peers.
Outbound Dial-Peer Matchinq Ordi
router(config]#
dial-peer hunt hunt-order-number
Specifies the hunt selection order fordial peers
Hunt-order- Description
number
Longest match inphone number, explicitpreference, random
selectiontne default hunt order number
Longest match in phone number, explicit preference, least recent use
Explicit preference, longest malch in phone number, random selection
Explicit preference, longest match in phone number, least recent use
Least recent use, longest match in phone number, explicit preference
Least recent use. explicit preference, longest malch in phone number
Random selection
Least recent use
The dial-peer hunt command specifies the hunt selection order for outbound dial peers, when
multiple outbound dial peers provide a potential match. The defined selection order applies to
all tvpes of dial peers: VoIP. POTS, and Multimedia Mail over IP (MMolP). The available
options are described in the table.
Hunt Order
Number
Description
0 Longest match in phone number, explicit preference, random
selectionthe default hunt order number
1 Longest match in phone number, explicit preference, least recent use
2 Explicit preference, longest match in phone number, random selection
3 Explicit preference, longest match in phone number, least recent use
4 Least recent use, longest match in phone number, explicit preference
5 Least recent use, explicit preference, longest match in phone number
6 Random selection
7 Least recent use
4-102 Implementing Cisco Voice Communications and QoS (CVOICE) vS.O 2010 Cisco Systems. Inc
Best Practices
This subtopic provides recommendations for dial-peer configuration.
Best Practices
Default POTS dial peer with the direct-inward-dial attribute
Provide Cisco Unified Communications Manager redundancy
Primary Unified
1010 102
Secondary Unified
Communication s Manager
10 10 10.3
dlal-paac voiea 1 pota
Incovtng callad-rounbar
dlrct-inward-dil
vole claai MS3 1
h225 elBBUC tcp aatabllah 3
illil-pHi volea 100 voip
pEiltnjiQ* 1
d.tintlon-pttrii 1. ..
....ion tarjat ip.4.10 .10 . 10.2
voka-clui h323 1
dial-paar vole. 101 voip
praEaranc* 2
d a at 1 rut ion-pat tarn 1. . -
ten taigat ipv*:io.io.10.3
volca-claaa h223 1
iPSTffe
Pfimsry united
Communroafions Managar
Server
Co rrimtifi ica Son s Manager
Server
Best practices should befollowed when configuring dial peers ona gateway. Toensure that
incoming PSTN calls aredirectly routed totheir destination based onthe called number
information, createa default POTSdial peer withthe direct-inward-dial attribute.
Note This should be the first POTSdial peer that youconfigureon the gateway. It should be the
onlydial peer that containsa period (.} forthe destination pattern and directinward dial. It
should not contain a port number.
Intheexample, dial peer 1is used toroutecallsaccording totheircalled number.
When configuring a gateway for interoperability witha CiscoUnified Communications
Manager cluster, provide redundancy byconfiguring at leasttwoVoIP dial peerswith thesame
destination pattern pointing totwodifferent CiscoUnified Communications Manager servers.
Use thepreference attribute to select thepriority orderbetween primary andsecondary Cisco
Unified Communications Manager servers.
In the example, dial peers 100and 101 are usedto route calls to IheprimaryCisco Unified
Communications Manager cluster unless it has lostconneetivily. If thecluster lost connectivity,
they are then routedto the backup or secondary CiscoUnifiedCommunicalions Manager
cluster.
The voiceclass configuration is usedto set the H.225timeout to the minimum recommended
value of 3 seconds. This setting reduces the fallback delay to the secondary Cisco Unified
Communications Manager serverwhenthe primary server fails. Bydefault, the 11.225 ICP
timeout is 15 seconds.
Path Selection Strategies
This topic describes common path selection strategies that are deployed in enterprise VoIP
env ironments.
ion btra
PSTN Requirement
Site-code dialing
Toll bypass
Tail-end hop-off (TEHO)
Call muting and path selection for Intersite calls
Digit manipulation to support site-codedialing
Call routing and path selection to route intersite
calls over WAN links with PSTN fallback
Digit manipulation to route calls over the WAN or
PSTN
Can routing and path selection to route PSTN
callsover the cheapest possible path
Digit manipulation to support PSTN fallback
When remote sites are involved, different path selection strategies are required. Multisitedial
plans include all ofthe requirements of a single-site dial plan, as well as theserequirements:
Site-code dialing: A tvpical requirement is the support of site-code dialing. Site-code
dialing allows users to place an intersite call by dialing a site code thai is typically three to
four digits long, followed bv the actual extension ofthe remote-site user. Call routing and
path selection can support this by using digit manipulationto prefix and strip off site codes
where necessary.
Toll bypass: foil bypass uses the WAN link for eall routing to avoid PSTNcharges for
intersite calls. This includes call routing and path selection for the actual call-routing
process, including fallback PS'I'N routing in case the WAN link fails. Digit manipulation is
also required to ensure proper number formatting.
Tail-end hop-off (TEHO): TF.HO is similar to toll bypass but extends the WANusage for
PSTN calls as well. The PSTN breakout should be as close as possible to the final PSTN
destination to decrease phone charges. The same requirements exist as with toll bypass.
4-104 Implementing Cisco Voice Communications and QoS |CVOICE| v8 0 2010 Cisco Systems, inc
Site-Code Dialing and Toll Bypass
This topic describes site-code dialing and toll bypass principles.
Requirements
Used primarily to solve overlapping numberingplan issues.
Users dial <intersite prefix> + <site code> + <user extension>
to reach a user in a specific site.
- <Site prefix> = <intersite prefix> + <site code>
For example: 802 = 8 and 02
Site codes may have different lengths.
Useful for few large and many small sites
Ambiguity must be avoided
The calling number should also includethe site code of the
calling party.
- This can be done via digit manipulation.
You may use site-code dialing tosolve issues with overlapping numbering plans. Because all
extensions of a site areprefixed with a unique sitecode, anoverlapping numbering plan (where
extensions inmultiple sitesoverlap) canbeturned intoa unique numbering plan.
When vou use site-code dialing, each site isassigned a unique site code. An intersite prefix
verifiesthat site codes are dialed. For example, a networkwiththreesites couldhave the
intersite prefix 8 and site codes 01, 02, and 03, resulting insite prefixes 801, 802, and 803. If a
user wants toplace a call to a remote siteuser, thedialed number would bethe intersite prefix
(8). followed by site code (forexample, 01)followed bytheactual extension. This form of
abbreviated dialing greatly improves theend-user experience byproviding shorter dialable
numbers.
Ihe calling-party number must include theappropriate sitecode. Thisallows called users to
call backthecalling party directly using theirmissed call list andreceived call list. Youcanuse
digit manipulation to support this as well.
Site-Code Dialing and Toll Bypass Example
This subtopic prov ides anexample of site-code dialing and toll bypass.
Necessary site prefix manipulation.
Prepend source-site prefixin callingnumber (originating gateway)
Stnpdestmation-site prefix incalled number(originating or terminating gateway)
Site A
Site Ptefix 801
destination-pattern SQ2.,,.
session target ipv4:ID.10.0.2 IP WAN
SiteB
Site Prefix 802
destination-pattern 801....
session target ipv4:10,10.0.1
1he figure shows a sample scenario for site-code dialing with toll bypass. Theenterprise uses 8
as the intersite prefix. Site A uses site code 01. Site B uses site code 02. Intersite calls arc
processed in the following order:
Step 1 A user in site B places a cal! from extension 2002 to extension 2001 in site A. fhe
user must prepend the intersite prefix and the site code of site A to the actual
extension, and therefore dials 8012001. When the originatinggateway sends the call
to the tenninatinggateway, it modifies the callingnumberby prepending its site
prefix. Such a modified calling number allows callback.
Step2 The call is routed over the IP WANlink to site A. The destination phone rings and
displavs the calling number 8022002that is, the site code 802 of site B followed
by the internal extension 2002.
Note The intersite prefix must be stripped fromthe called number for successful routing to the
internal extension This stripping can be configured either on the originating gateway or
terminating gateway. Stripping on the originating gateway can be implemented using the
outgoing translation profile that is attached to the outbound VoIP dial peer. Stripping on the
terminating gateway can be best achieved by an incomingtranslation profilethat is applied
to a source group. A source group can use an access list to specify the sources from which
the calls should be subjected to the incoming translation profile.
4-106 Implementing Cisco Voice Communications and OoS (CVOICE) vS.O 2010 Cisco Systems. Inc.
Site-Code Dialing and Toll Bypass with Backup Example
This subtopic provides an example of site code dialing and toll bypass with PSTN backup.
Site-Code Dialing and Toll Bypass wi
Backup Example
Site A
Site Prefix1 S01
DID; 2005552xxx
WAN is the preferred
patti wth preference 0.
seHoioa-target ipv4:10.10.0.1
prefix 1200555
preference 1
port 0/0/0:23
The figure shows an example of site-code dialing with PSTN backup. The PSTN is used as the
redundant path for intersite calls ifthe IP WAN fails. Intersite calls are processed in this order:
Auser in site Bdials 8012001to reachan extension in site A. Theoriginating
gateway matches the VoIP dial peer 801 as the best dial peer for the call, but the
WANis unavailable so the redundant POTS dial peer 22 is used instead. Digit
manipulation that isconfigured on that dial peer prepends the required PSTN prefix
to the called number and the call is sent into the PSTN.
The call is routed over the PSTNto site A. The terminating gateway adjusts the
called and calling numbers and delivers the call totheintended recipient.
Stept
Step 2
Note
Digit manipulation that isrequired in this scenario isdescribed later in the lesson.
Configuring Site-Code Dialing and Toll Bypass
fhis topic describes how to configure site-code dialing and toll bypass.
i Configure voice translation rules and profiles for VoIP
intersite routing.
'2 Define dial peers for VoIP intersite routing.
:> Configure voicetranslation rules and profiles for PSTN
backup routing.
^ Define dial peers for PSTNintersite routing.
Follow thesestepsto configure site-code dialingand toll bypass:
Step 1 Configure voice translationrules and profiles for inbound and outbound VoIP
intersite routing.
Step 2 Define thedial peers for VoIPintersite routingthat routethe call over the WAN.
Step 3 Configure voicetranslation rules and voicetranslation profiles for inbound and
outbound PSTN intersite routing(PSTN backup).
Step4 Define the dial peers for POTS intersite routingthat routethe call usingthe PS'I'N.
Step 1: Configure Voice Translation Rules and Profiles for VoIP
Intersite Routing
This subtopic explains the firsi step in site-code dialing and toll bypass configuration, in which
the VoIP translation rulesandprofiles aredefined.
Step 1: Configure Voice Translation Rules
and Profiles for VoIP Intersite Routing
Site A
Site Prefix: 801
Ext: 2xxx
IPWAItf
SiteB
Site Prefix: 802
Ext: 2xxx
tula 1 /*!/ /B012/
rule 1 /"8012/ /2/
voice translation-profile inter-out
translate calling 1
voice translation-profile inter-in
translate called 2
voioo translation-rule 1
rule 1 /'2/ /8022/
rule 1 /*8022/ nl
voice translation-prof ile inter-out
translate calling 1
voice translation-profile inter-in
translate called 2
Follow these steps at each site to configure translation rules and profiles for WAN routing:
Step 1 Create a rule that prefixes the site code tothe calling number.
Step 2 Create arule that strips offthe site code from the called number.
Step 3 Create avoice translation profile to prefix the site code to the outbound calling-party
number.
Step 4 Create avoice translation profile to strip off the site code from the inbound called-
party number.
Step 2: Define Dial Peers for VoIP Intersite Routing
This subtopic explains the second step in site-code dialing and toll bypass configuration,
which the VoIP dial peers are defined.
Step 2: Define Dial Peers for
intersite Routing
Site A
Site Prefix: 801
Ext: 2xxx IP WAN
SiteB
Site Prefix: 802
Ext: 2xxx
When the \ oice translation profiles for VoIP routing aredefined, the VoIP dial peers for
intersite routing \ ia the WAN must beconfigured. In this example, thecalled numbers are
modified on theterminating gateways in theinbound dial peers.
4-110 Implementing Cisco Vorce Communications and QoS (CVOICE) v8.0
Step 3: Configure Voice Translation Rules and Profiles for
PSTN Intersite Routing
This subtopic explains the third step in site-code dialing and toll bypass configuration, in which
the PSTN translation rules and profilesare defined.
Step 3: Configure VoiceTranslation Rules
and Profiles for PSTN Intersite Routing
Site A
Site Prefix: 801
Ext: 2xxx
IP WAN
SiteB
Site Prefix: 802
Ext: 2xxx
rule 1 !"2/ /2005552/
rule 1 /"8022/ /13005552/
voice translation-profile 802PSTN
translate calling 3
translate called 4
rule 1 /"2/ /3005552/
rule 1 /"S012/ /12005552/
translate calling 3
tran slate called 4
To support PSTN fallback routing if the WAN link fails, two additional voice translation rules
and oneprofile must bedefined foreach site. This manipulation adjusts thecalling and called
numbers to the public numberingscheme used in the PSTN.
Note Thesetting ofthe calling party numberis not mandatory because the telcosets the calling
number when the call enters the PSTN network. It is recommended in situations when a
range of DID numbers is available to select the appropriate number for callback.
Step4: Define Dial Peers for PSTN Intersite Routing
This subtopic explains the fourth step in site-code dialing and toll bypass configuration, in
which the PSTNdial peers are defined.
Define Dial Peers
ng Intersite Routi
Site A
Site Prefix 801
Ext: 2xxx
destination-pattern SO22...
port 0/0/0:23
preference 1
translation-profile outgoing
IP WAN
Site B
Site Prefix: 802
Ext: 2xxx
dial -peer voice 8012 pots
dea tinatio -pattern 8012.
por t 0/0/0 23
preference 1
tra nslatio -profile outgoing 801PSTN
Final!;. the PSTN translation profile is applied totheoutbound dial peers pointing tothe PSTN
network. The destination pattern of those peers matches the called number, because the users
dial the intersite prefixand thesite codeto place intersite calls independently ofthe path that
thecall takes. The POTS dial peers areconfigured with a worse preference, that is. numerically
higher, than the default value of 0.
4-112 Implementing Cisco Voice Communications and QoS (CVOICE) v8 0 )2O10 Cisco Systems. Inc
Outbound Site-Code Dialing Example
fhis subtopic explains the site-code dialing operations for outbound calls.
Outbound Site-Code Dialing Example
rule 1 /'2/ /8012/
voice translation-profile inter-out
translate calling 1
session-target ipv*;10.10.0.2
translation-profile outgoing inter-
Site Prefix 801
A DID 200-555-xxxx
rule 1 /'2/ /200SSS2/
rule 1 /*8022/ /13005552/
translate calling 3
translate called 4
destination-pattern B022...
preference 1
port 0/0/0:23
translation-profile outgoing 802PSTN
incoming Outgoing
Called number 8022001 . 8022001
Calling number 2001 8012001
Site Prefi a- 802
. { DID: 300-555-xxxx
IP-WAI^ ; 10.10.0.2
Incoming outgoing
Called number B022001 13005552001
Callingnumber 2001 2005552001
The figure summarizes theprocessing for outbound intersite callsas follows:
Step1 A userwith extension 2001 insiteAdials 8022001 toreachanendpoint in the
remote site. Tlie originating gateway receives a call withthe called number 8022001
andthecalling number 2001. Thecalled number matches twodial peers: 802and
8022. Dial peer802is matched because it has thebest preference, andthetranslation
profile inter-out is applied totheoutbound call. Thus, thecall is routed tothe
remote site with the called number 8022001 (unchanged) and the calling number
8012001 (modified by the translation rule).
Step2 If theWAN fails, thecall will be routed using dial peer8021 withpreference1. Ihe
translation-profile 802PSTN is used, whichmodifiesthe callingnumber to
12005552001 and the called number to 1-300-555-2001that is, the call can be
routedby the PSTN to the remotesite.
Note The calling number for the PSTNpath can be adjusted inthe outgoingtranslation profile that
is attached to the voice port instead of the setting inthe outbound POTS dial peer.
Inbound Site-Code Dialing Example
This subtopic explains site-code dialing operations for inbound calls.
inbound Site-Code Dialing Exampl
S*te Prefix 801
DID 20Q-555OUHX
,3-001 \9-WAH
Called number 8022001 2001
Calling number 8012001 6O1Z001
Sile Prefi;, 802
DID. 300-555-xxxx
101002
ce tranelation-rula 2
le 1 /'B022/ 111
ce translation-profile inter-ir
analate called 2
1-peer voice 8011 voip
Btination-pattarn 8012...
ssion-target ipv4:10.10.0,1
coming called-number 802
anslation-profile incoming inte
The figure summarizes theprocessing for inbound intersite calls. Thisexample illustrates the
option thattransmits the sitecode in thecalled number, andthesileprefix is stripped on the
terminating gateway in the inbound dial peer.
Tail-End Hop-Off
This topic describes the characteristics oftail-end hop-off (TEHO).
Tail-End Hop-Off
- Extends the concept of toll bypass
Uses the WANfor PSTN calls as much as possible
Uses PSTN breakouts closest to the final PSTN destination
Uses PSTN paths as possible backup
TEHO extends theconcept of toll bypass. Instead of routing only intersite calls overanIP
WAN link. TEHOalso usesdie IP WANlink for PSTN calls, 'fhe goal is to route a call
through the IP WAN as far as possible and break out to the PSTN on the gateway nearest to the
destination. Aswith toll bypass, PSTN fallback should always bepossible incase theII' WAN
link fails.
Caution Some countries do nol allowTEHO. When implementingTEHO, ensure that the deployment
complies with national legal requirements.
TEHO Scenario
This subtopic provides asample scenario that illustrates a typical TEHO implementation.
TEHO Scenario
Site A gateway is
used as the PSTN
breakout
7^
SiteAmAieaA
200555
Call is routed lo site A via tha WAN
IP WAN
The figure illustrates call forwarding in a TEHO scenario, as follows:
Stepl User in site 13 dials 9-1-200-555-1 111 to reach a PS'I'N subscriber in the same area
as site A.
Step 2 Ihe eall is routed to site A using the IP WAN link.
Step 3 Tlie gatewax in site A fonvards the call as a local call to the PSTN subscriber.
Step 4 The PSTN phone rings.
4-116 Implementing Cisco Voice Communications and QoS (CVOICE) v8 0 2010 Cisco Systems, inc
Configuring TEHO
This topicdescribes howlo implement TEHOon Ciscovoicegateways.
TEHO Configuration Overview
Define the VolP outbound digit manipulation for TEHO.
Define the outbound VoIP dial peerforTEHO.
Define the outbound POTS dial peer for TEHO.
Follow these steps to configure TEHO functionality:
Step 1 Define the VoIP outbound digit manipulation.
Step 2 Define the outbound VoIP dial peer.
Step 3 Define the outbound POTS dial peer.
TEHO Configuration Example
Thissublopic pro\ idesa sample configuration that illustrates a typical TEHOimplementation.
iqu
IP WAN
Site B in Area B -r
\ \piD 300555xxx- ^
\ \ 101002
V
R2
translation-rule 10
1 /'2/ /13005S52/
translation-profile pstn-out
late calling 10
1-peer voice 912001 voip
session-target ipv4:10.10.0.1
translation-profile outgoing pstn-out
The figure illustrates the eon figuration ofthe 1EHO components:
1. A translation profile is needed lo adjust the calling number ofthe caller to the PSfN
numbering scheme. I he calling number includes the DID prefix ofthe originating site.
2. A VoIP dial peer is defined to match the PS I N numbers that should be redirected via the
VoIP path to the remote break-out gateway, fhis dial peer has the outgoing translation
profile attached that sets the calling number to the appropriate PSTN number.
3. A POTS dial peer is configured on the break-out gateway that matches the called number
of long-distance PS'I'N calls and forwards the TEHO calls to the PSTN while keeping the
calling number that as set by the originating gateway.
Note Depending on the enterprise policy, the calling number for the TEHO calls could be set to
the DID number of the break-out gateway This approach is less common, however,
because it creates potential issues with the return path for callback.
Summary
This topic summarizes thekey points thatwere discussed inthis lesson.
Summary
The call routing logicof Cisco voicegateways is based on dial peer
matching.
Routers must match the inbound and outbound dial peers to
successfully complete a call.
The common path selection strategies of a multi-site environment
are site-code dialing, toll bypass, and TEHO.
Site-code dialing uses the concept of prefixing remote extensions
with a site code and can be combined with toil bypass to route calls
over the WAN instead of the PSTN.
Site-code configuration requires that each site be assigned a unique
site code.
TEHOextends the concept of toll bypass by routingcalls over the
WAN to the closest PSTN breakout.
TEHO configuration requires that all calls be routed over the
WAN unless the WAN is down or has been oversubscribed.
Lesson 51
Configuring Calling Privileges
Overview
Calling privileges on Cisco IOS gateways are dial plan components that define the types of
calls that aphone, or group of phones, is able to place. This lesson describes the concept ot
calling privileges and how they can be implemented on Cisco IOS gateways using class of
restriction (COR).
Objectives
Upon completing this lesson, you will be able to describe how to configure calling privileges in
agateway. This ability includes being able to meet these objectives:
Describe calling privileges characteristics and explain their operations
Describe how toimplement calling privileges onCisco IOS gateways
Describe how to implement calling privileges in Cisco Unified SRST and Cisco United
Communications Manager Express and how it differs from the implementation on voice
gateways
Describe how to configure COR
Describe how to verify COR
Calling Privileges Characteristics
This topic describes the concept ofcalling privileges.
Calling Privileges
Characteristics ofcalling privileges:
Define the destination that a user is allowed to dial
Implemented on Cisco IOS gateways using COR
Relyon proper call routing
Require high dial-peer granularity
Different destinations defined inseparate dial peers
Calling privileges define the destination to which auser is allowed to dial and connect. COR is
aCisco voice gateway feature that enables class of service (CoS) or calling privileges to be
assigned to users. Calling privileges are implemented on Cisco IOS gateways using COR. COR
is most commonly used with Cisco Unified Communications Manager Express, hut can be
applied to any dial peer.
COR is dependent on properly configured call routing. Ifyour calling privileges are not being
enforced effectively, the problem may be linked lo an incorrect call routing configuration.
COR requires agranular dial-peer configuration. The common 9T destination pattern cannot be
used as a"catch all" outgoing dial peer to the PSTN. More specifie dial peers are necessary to
distinguish between internal calls, local calls, long-distance calls, international calls, and
services suchas emergency 911.
mm
r-
fa^^
Dial-Peer Granularity
This subtopicdescribes the requirement for dial-peergranularity whendeploying calling
privileges.
Dial-Peer Granularity
niemational Calls (Variable Length)]
dial-peer voice 911 jots
forward-digito all
port 0/0/0:23
dill-peer voice 9311 pots
dee tioatioa-pattern 9911
forward-oigl te 3
port 0/0/0:23
dsetioation-pattarn 912-91
port 0/0/0:23
deetiDation-pattern 912-9] .. [2-9]
port 0/0/0:23
dee tiost ion-pattern 91[2-9]..[2-9]
pretix 1
port 0/0/0:23
prefix 011
port 0/0/0:23
The figure illustrates dial-peer granularity when deploying calling privileges. The dial plan
consists of multiple public switched telephone network (PSTN) dial peers that establish a
baseline for COR settings. This example is specific to the North American Numbering Plan
(NANP) and will be constructeddifferently in other parts ofthe world.
The 911 dial peer is used for emergency calls to ihe PSTN. The forward-digits all command
sends all matched digits (911, in this case) to the PSTN. Without this command, the dial peer
would be matched, but no digits would be sent to the PSTNbecause ofthe default digit-strip
command.
The 9911 dial peer is also used for emergency calls, but this time it includes the PSTN access
code 9. Only three digits are sent to the PSTNas a result ofthe forward-digits 3 command.
The PSTNaccess code 9 musl not be included in the call setup.
The 97 dial peer is used for PSTNlocal calls for seven-digit dialing in the United Stales.
"fhe910 dial peer is used for PSTNlocal calls fbr 10-digit dialing in the United States.
fhe 9110 dial peer is used for PSTNnational or long-distance calls for 1l-digit dialing in the
United States. Because the exactly matched digits are 91, the national identifier 1 needs to be
prefixed. This is done using the prefix 1 command.
The 9011 dial peer is used for PSTNvariable-length international calls fromthe United States.
Because 9011 will be stripped due to Ihe digit-strip setting, the prefix 011 command is used to
prefix the correct international identifier to the called number.
Implementing Calling Privileges on Gateways
This topic describes how to implement calling privileges on voice gateways.
Calling Privileges on Gateways
* Calling privileges are implemented using COR.
COR lists contain CORs and are used to control call routing.
COR lists are assigned to dial peers:
Incoming COR list
Outgoing COR list
* For each call, the incoming COR list is matched against the
outgoing COR list;
If the outgoing COR list is a subset of the incoming COR
list, the call is routed.
If no incoming COR list is configured, the call is always
routed.
The COR feature provides the ability lo deny certain eall attempts based on the incoming and
outgoing CORs pro\ isioned on the dial peers.
The COR is used to speeifv which incoming dial peer can use which outgoing dial peer Lo make
a eall. Each dial peer can be provisioned with an incoming and an outgoing COR list. COR
functionalilv provides the abilitv to deny certain call attempts based on the incoming and
outgoingCORs that are prov isionedon the dial peers. This functionalityprovides flexibility in
network design, allows users to block calls (for example, calls lo 900 numbers), and applies
different restrictions to call attempts from different originators.
Ihe fundamental mechanism at the center ofthe COR functionality relies on the definition of
incoming and outgoing COR lists. F.ach COR list is defined lo include a number ol'members,
which are simpK tags previously defined within Cisco IOS Software. Multiple CORs are
defined, and COR lists are configured that contain these CORs. F.ach COR list is then assigned
to dial peers as an incoming or outgoing list using the corlist incoming or eorlist outgoing
command.
When a call goes through the router, an incoming dial peer and an outgoing dial peer are
selected based on the Cisco IOS dial-peer routing logic. If COR lists are associated with the
selected dial peers, the followingadditional check is performed before extending the call:
If the CORapplied on an incoming dial peer (for incomingcalls) is a superset or equal to
the CORapplied to the outgoing dial peer (for outgoing calls), the call goes through.
If the COR applied on an incoming dial peer (for ineomingcalls) is not a superset or equal
lo the COR applied to the outgoingdial peer (for outgoing calls), the eall is rejected.
___*
Note Incoming andoutgoingare termsthatare usedwith respect tothe voice ports. For
example, ifyou hook upa phone thatis connected toone ofthe Foreign Exchange Station
(FXS) ports ofthe router andtry to make a call from that phone, itis an incoming call for the
router and voice port. Similady, ifyou make a call to that FXSphone, then it is an outgoing
call.
If noCORlist statements areapplied tosomedial peers, thefollowing properties apply:
When noincoming eorlistcommand is configured ona dial peer, thedefault incoming
CORlist is used. Thedefault incoming CORlist has the highestpossiblepriority, and it
therefore allows thisdial peerto access all otherdial peers, regardless of theiroutgoing
COR list.
W;hen no outgoingcorlist command is configured on a dial peer, thedefaultoutgoingCOR
list is used, 'fhe defaultoutgoingCORlist has the lowestpossible priority, and it tlierefore
allowsall other dial peers to accessthis dial peer, regardless of their incoming CORlist.
>2010 Cisco Systems. Inc. Dial Plan Implementation 4-125
COR Elements
This subtopic explains the logical elements ofthe COR implementation.
1 Call 100
dials 1XXX
2 Call 200
dials 2XXX
Outgoing COR Nst is
NOT a subset gf
incoming COR lot
Outgoing CORm~
a subset of incoming
COR list
dial'poer vole* 3 pots
destination-pattern 1~
The figure illustrates the logical elements of COR. In this example, the VoIP dial peer is
associated with the el incoming COR list, with members A. B, and C. You can dunk of
members ofthe incoming COR list as "keys."
The first plain old telephone service (POTS) dial peer has a destination-pattern of I... and is
associated with the c2 outgoing COR list, with members A and B. The second POTS dial peer
has a destination-pattern of 2... and is associated with the c3 outgoing COR list, with members
A. B. and D. You can think of members ofthe outgoing COR list as "locks."
For the call to succeed, the incoming COR list ofthe incoming dial peer must have all the keys
needed to open all the locks of the outgoing COR list ofthe outgoing dial peer.
In the example in the figure, a first VoIP eall with deslinalion 100 is received by the router. The
Cisco IOS call routing logic matches the incoming eall leg with the VoIP dial peer and the
outgoing call leg with the first POTS dial peer. The COR logic is then applied. The el
incoming COR list has all the keys that are needed for the c2 outgoing COR list locks (A and
B). so the call succeeds.
A second VoIP call with destination 200 is then received by the router, fhe Cisco IOS call
routing logic matches the incoming call leg with the VoIP dial peer and the outgoing call leg
with the second POTS dial peer. The COR logic is then applied; because the el incoming COR
list is missing one kev for the c3 outgoing COR list (D). the eall is rejected.
4-126 Implementing Cisco Voice Communications and QoS (CVOICE! v8.0 2010 Cisco Systems, Inc.
COR Logic
This subtopicexplainsthe logicofthe callingprivileges implementation.
COR Logic
Scorning COR Ust
CORHtx*catt*
911
Local
LD
NTL
CORWtOto
911
Local
Outgoing COR List
ct
Calling privileges on Cisco IOS gateways use two components:
COR name: A COR name (also called a COR label) is the building block of calling
privileges. An individual COR name is actually a member of a list that is known as a COR
list.
COR list: A COR list contains multiple COR names and is bound to dial peers.
When a call is routed, the gateway checks the COR list ofthe inbound dial peer and the COR
list ofthe outbound dial peer. The Call Routing with COR Lists table describes the various
results depending on the configuration:
Call Routing with COR Lists
COR List on
Incoming Dial
Peer
COR List on
Outgoing Dial
Peer
Result Reason
No COR No COR Call
succeeds
COR is not in the picture.
No COR The COR list
applied for
outgoing calls
Call
succeeds
The incoming dial peer, by default, has the highest
COR prioritywhen no COR is applied. Therefore, ifyou
apply no COR for an incoming cal! leg to a dial peer,
then this dial peer can make calls out of any other dial
peer, regardless of the COR configuration on the
outgoing dial peer.
COR List on COR List on Result Reason
incoming Dial Outgoing Dial
Peer Peer
The COR list No COR Call The outgoing dial peer, by default, has the lowest
applied for succeeds priority. Because there are some COR configurations
incoming calls for incoming calls on the incoming, originating dial
peer, it is a superset of the outgoing call COR
configurations on outgoing, terminating dial peer.
The COR list The COR list Call The COR list for incoming calls on the incoming dial
applied for applied for succeeds peer is a superset of COR lists for outgoing calls on the
incoming calls outgoing calls outgoing dial peer.
(Superset of (Subset of
COR lists COR lists
applied for applied for
outgoing calls incoming calls
on the on the
outgoing dial incoming dial
peer) peer)
The COR list The COR list Call cannot COR lists for incoming calls on the incoming dial peer
applied for applied for be are not a superset of COR lists for outgoing calls on
incoming calls outgoing calls completed the outgoing dial peer
(Subset of (Superset of using this
COR lists COR lists outgoing
applied for applied for dial peer
outgoing calls incoming calls
on the on the
outgoing dial incoming dial
peer) peer)
4-128 Implementing Cisco Voice Communications and QoS (CVOICE] vS.O
COR Implementation Example
This subtopic provides a COR implementationexample.
COR Implementation Example
Outgoing COR Lot
Atypical COR implementation defines a COR name foreach number of anoutgoing dial peer,
then defines a listthat contains only that COR name and assigns thatlistascorlistoutgoing for
this outgoingdial peer. For example, the dial peer withdestination pattern901 IT can have a
corlist outgoing that contains CORINTL, as showninthe example.
Thisexample defines four COR names: 911. Local, Long Distance (LD), andInternational
(INTL). These four CORs are used tocreate three incoming CORliststhat will be assigned to
phones and users:
Lobby: This COR list contains the COR names 911 and Local. This list will allow users to
place emergency calls and local PSTN calls.
Employee: This COR list contains the COR names 911, Local, and LD. This COR list will
allowusersto placeemergency calls, local calls, and long-distance PSTN calls.
Executive: This CORlist contains the CORnames 911, Local, LD. and INTL. This COR
list will allow users to place any PSTN call.
ACORlist will be assigned to an outgoingPOTS dial peer for international calls:
INTLCall: This COR list contains the COR INTL.
When a eall is routed using the incoming COR listExecutive and the outgoing COR list
INTLCall. the call succeeds because COR name INTL is included in the COR list Executive.
When a call is routed using theincoming COR listEmployee and the outgoing COR list
INTLCall. the call isblocked because COR INTL isnot included inthe COR list Employee.
Implementing Calling Privileges on SRST and
This topic describes howto implement calling privileges on Cisco Unified Survivable Remote
Site Telephonv (SRST) and Cisco Unified Communications Manager Express.
Calling Privileges on SRST and Cisco
Unified Communications Manager
Express
<
Cisco Unified Communications Manager Express and Cisco
Unified SRST use the standard Cisco IOS COR concept.
Can be configured for inbound and outbound direction:
Inbound: Restrict destinations to which a user can dial
Outbound: Restrict who can call a user
For Cisco Unified Communications Manager Express, COR
lists are assigned to ephones.
For Cisco Unified SRST, COR lists are assigned to number
ranges.
When implementing COR inCisco Unified SRST andCisco Unified Communications Manager
Express, the virtual dial peers for the registered endpoints are created dynamically by the
router. Thus, theCORlists are appliedto theephone-dns (for Skinny ClientControl Protocol
[SCCP] endpoints) and tovoice register pools (forSession Initiation Protocol [SIP] endpoints)
instead of applying them to virtual dial peers.
For Cisco UnifiedCommunications Manager Express, the COR list is directly assigned to the
appropriate endpoint (SIP) or directory number (SCCP) andwill be included inthevirtual dial
peer. Both inbound and outbound COR lists can beapplied toSCCP and SIP endpoints. An
inbound COR list restricts the dialable destinations, whereas an outbound COR list delines who
can reach the endpoint.
For Cisco Unified SRST. the endpoints are not statically configured on the Cisco IOS gateway.
Instead, the gateway dynamically creates the endpoints that re-home asa result of lost
connectivitv to the Cisco Unified Communications Manager. To assign a COR list in SRST
mode, a CORlist is appliedto a rangeof directory numbers in global SRSTconfiguration
mode.
Note COR is not limited to Cisco Unified Communications Manager Express or Cisco Unified
SRST. CORcan be appliedto any inbound and outbounddial peer on a Cisco IOSgateway.
4-130 Implementing Cisco VoiceCommunications and QoS (CVOICE] v8.0
J2010 Cisco Systems, Inc.
Calling Privileges Implementation at a Glance
This subtopic provides an overview ofcalling privileges implementation ona Cisco Unified
SRST and Cisco Unified Communications Manager Express.
Calling Privileges Implementation at a
Glance
Manager Express
register pool 1
incoming Executive default
ephone-dn 1
corliet incoming Bxecuti
Incoming COR List
1
COW Lwt EweutiW
9'-'.
Lccai
,.D
INTL
SRST ***%?
f
call-manager-fallback
cor incoming Executive 1 200 0 2100
I
Outgoing COR List
COR List INTLCal
This Cisco UnifiedCommunications Manager Express configuration assigns the incoming
COR list Executive to voice register pool 1 (SIP endpoint) and ephone-dn 1 (SCCP endpoint):
voice register pool 1
cor incoming Executive default
i
ephone-dn 1
corlist incoming Executive
This SRST configuration assigns the incoming COR list Executive to all phones with the
director)' number 2000 to 2010:
cor incoming Executive 1 2000 - 2010
Note The number that precedes the directory number range in the SRST configuration is the COR
list tag. Up to 20 tags can be configuredthat is, up to 20 different COR lists can be used
for Cisco Unified SRST phones.
i 2010 Cisco Systems, Inc Dial Plan Implementation 4-131
Configuring COR
fhis topic describes howto configure calling privileges on a voice gateway running Cisco
Unified Communications Manager Express.
COR Configuration Overview
When using a standalone gateway:
1 Define COR labels
'/. Configure COR lists
- Incoming
Outgoing
3 Assign COR lists to dial peers
When using Cisco Unified Communications Manager
Express:
' Assign COR lists to endpoints
SCCP: ephone-dn
SIP: voice register pool
Follow these steps to configure call privileges on a voice gateway:
Step 1 Define the four individual tags (COR names) lo be used as COR list members with
the command dial-peer cor custom.
Step 2 Define the COR lists that will be assigned as ineoming or outgoing to the dial peers
with the command dial-peer cor list corlist-name.
Step 3 Associate COR lists with existing VoIP or POTS PSTN dial peers by using the
command corlist {incoming | outgoing} corlist-name within the dial-peer.
Step 4 When deploving COR functionalit\ on Cisco Unified Communications Manager
Express, apply COR lists directly to the endpoints.
Configuring COR
This subtopic explains how to configure the two basic COR elementsCOR names and COR
lists.
Configuring COR
router(conflg)#
dial-pear cor custom
Enters the mode to define COR labels (names)
router(config-dp-cor)#
1
name class-name
Defines the name (label) for a custom COR
rou ter (con f iq) tt
dial-peer cor list list-name
- Creates a COR list
router(eon(ig-dp-corlist)#
member
Adds a name (label) as memberto a COR list
Thedial-peercor custom command enterstheconfiguration modetodefine CORnames
(labels).
The name command in dial-peercor customconfiguration modedefines the namesof
capabilities. This definition isnecessary' tospecify COR rules and apply them tospecific dial
peers. Amaximum of 64 CORnamescan be configured on a gateway.
The dial-peercor list command defines a CORlist name. ACOR listspecifies a capability set
that is used in COR checkingbetween incoming and outgoing dial peers.
The member command in the dial-peer eor list configuration mode adds a COR name as
member to a dial peer COR list.
Assigning COR
This subtopic explains how to assign dial privileges lodial peers and Cisco Unified IP phones.
Assigning COR
router(config-ephone-dn)#
I corlist {incoming | outgoing} cor-list.
Applies COR list to dial peer or ephone-dn in incomingor outgoing
direction (dial peers and ephone-dns)
router(config-register-pooII#
cor {incoming outgoing} cor-list-name {cor-list-number
starting-number [- ending-number] }
Applies COR list to SIP endpoint in incoming or outgoing direction
List-based mechanism assigns COR parameters to specific set of
directory number ranges (up to 10 directory numbers configured per
endpoint)
cor-list-number\s the identifier
starting-number to ending-number defines the range (can be 1)
Thecorlist incoming and corlist outgoing commands arc usedto applyCORlists inthe
ineomingor outgoingdirection to SCCPendpoints and dial peers lhat are not associated with a
registered endpoint. fhe command is applied either in ephone-dn configuration mode (for
SCCPendpoints) or in dial-peerconfiguration mode(dial peersnot associated withendpoints).
Thecor command invoiceregisterpool configuration modeassignsa CORlist in the incoming
or outgoing direction to SIPendpoints. A Hst-based mechanism assigns CORparameters to a
specific set of number ranges, which allows a different COR definition for various numbers that
are associated with the endpoint.
4-134 Implementing Cisco Voice Communications and QoS (CVOICE) v8.0 '2010 Cisco Systems, Inc.
Configuring COR Labels and Lists (Steps 1 and 2)
This subtopic provides asample configuration for the first two steps that cover the definition of
CORnames (labels) and lists.
Configuring COR Labels and Lists
(Steps 1 and 2)
1. Define COR labels
dial-peer cor cu stom
name 911
name local
name id
nans intl
MM ac
2a. Configure incoming COR lists
dial-peer cor liflt lobby
member 911
member local
dial-peer cor list sales
member 311
member local
member Id
member exec
dial-peer cor list executive
member 911
member local
member Id
member intl
member exec
2b. Configure outgoing COR lists
dial-peer cor list giicall
member 911
dial-peer eor list localcall
member local
dial-peer cor list Ideal1
member Id
dial-peer eor list intlcall
member iatl
dial-peer cor list execcall
member exec
The figure provides a sample configuration for the first iwo steps that cover Ihe definition of
CORnames (labels) and lists. The names (labels) are defined in Step 1 using the dial-peer cor
custom command. The incoming CORlists are configured in Step 2a, and the outgoingCOR
lists in Step 2b. usingthe dial-peer cor list command.
Assigning COR Lists to Dial Peers (Step 3)
This subtopic prov idesa sample configuration for the thirdstep, inwhichthe CORlistsare
attached to dial peers.
Assiqninq COR Lis!
Gaetvay
dial-peer v oice 911 pota
deatination-p ttern 911
f orward-d Oit all
corlist ou tgoi g 91 leal 1
port 0/0/0 :23
dial-peer v oice 9911 pots
destinatic n-pa ttern 9911
f orwa rd - d. qit 3
corlist ou tgoi ng 91 leal 1
port 0/0/0 :23
dial-peer v oice 9 po tB
destination-pa ttern 912-91
eorlist ou tqoi ng lo =alcall
port 0/0/0 -.23
dial-peer v oice 91 p JtS
deatination-pa ttern 91 [2-91 . . [2- 91
prefix 1
corlist ou tgoi ng Id ;all
port 0/0/0 r23
dial-peer v oice 9011 pots
destination-pa ttern 9011T
prefix 011
corlist ou tgoi ng in lcall
port 0/0/0 -.23
1he figure illustrates how to attachoutgoingCORlists to dial peers. This example showsonly
outbound POTS dial peers, but COR lists can also be applied to inbound dial peers using the
corlist incoming command. Other dial peer tvpes. such as VoIP or Multimedia Mail over IP
(MMolP) can also be configured with incoming or outgoing COR lists.
In this scenario, no COR lists are applied to inbound POTSdial peers in the incomingdirection.
Therefore, ineoming PSTNcalls will be forwardedto their respective destinations without any
restrictions.
4-136 Implementing Cisco Voice Communications and QoS (CVOICE] vS.O 2010 Cisco Systems Inc.
Assigning COR Lists to Endpoints (Step 4)
S . .- .: fXr 5tpn A in
CtouScdCon.mimi.-ion. Manager Express endpo.nl,
Assigning COR Lists to Endpoints (Step 4)
Sales (SIR
voce reaster pool 1
voce register areaory
number: 2
CiscounifiedCommunications
Manager Express
Lobby1003
voice register pool 1
cor incoming sales 1
2
ephone-dn 1
corlist incoming axe
eorlist outgoing e*a
cutive
ccall
port 0/0/0
corlist incoming lobby
Executive (SCCP)
Bptime-dri. 1
that does not include the EXEC label. _
.eiep^n^^
associated with them.
4-138
Configuring COR for SRST
This subtopic proves aconfiguration example for Cisco Unified S
Configuring COR for SRST
cor incoming executive 1 2000 - 2i(
cor outgoing exec-call 2 2000 - 210C
lied SRST.
To configure COR fbr Cisco Unified SRST
mode.
use the corlist command in SRST configuration
..noted 6rte;:; ^*::*rr-a- *^*-
I'nified Co,micaLS Managertl hIk rg12 """^"""^*'** Cis
implemenling Cisco Voice Communications and QoS (CVOICE) 8.i
>2010Cisco Systems. Ir
n^^^
Verifying COR
Thistopic describes how to verify the COR settings.
COR Verification Overview
show dta(-poer cor
show dial-peer voice
Descnptiofi
Displays the COR names (labels) and
lists
Displays the parameters of a voice dial
peer, including incoming and outgoing
COR list
show voice register pool Displays the parameters of a SIP
endpoint, including incoming and
outgoing COR list
show running-config | begin Views the configuration of SCCP
ephone-dn ephone-dns, including incomingand
outgoing COR list
You can verify COR settings with these four commands:
showdial-peer cor: This command displaysthe CORnames(labels)and lists applicable to
all types of dial peers and endpoints.
showdial peer voice: This command displaysthe parameters of a voice dial peer,
including incomingand outgoing COR list.
show voice register pool: This command displays tlie parameters of a SIP endpoint,
including incoming and outgoing COR list.
show running-config | begin ephone-dn: This command displays the configurationof
SCCPephone-dns, including incoming and outgoingCORlist. Because there is no SCCT-
specific command for displaying CORparameters, you must viewa part ofthe gateway
configuration.
) 2010 Cisco Systems, Inc. Dial Plan Implementation 4-139
Verifying COR Names and Lists
This subtopic explains how to view generic COR parametersCOR names and COR lists.
Verifying COR Names and List;
router# show dial-peer cor
Class of Restriction
name: 911
name: loca1
name: intl
COR list <executive>
member: 911
member: local
member: intl
COR list <lobby>
member: 911
COR list <911call>
member: 911
COR list <localcall>
member: local
COR list <intlcall>
member: Intl
fhe show dial-peer eor command display s the COR names (labels) and lists.
4-140 Implementing Cisco Voice Communications and QoS (CVOICE) vS.O i 2010 Cisco Systems. Inc
Verifying Dial Peer COR Settings
This sublopic explains how to view COR lists applied to dial peers.
Verifying Dial Peer COR Settings
router# show dial-pear voice 911
Voic eKncapPe er911
peer type = voice, system default peer = FALSE,
information type = voice,
description = ~',
tag - 911, destination-pattern = ~911',
Incoming COR litiwusim capability
outgoing COS XiattSlloaXl
The show dial-peer voice command displays the parameters of all or selected dia peers. The
parameters include the incoming and outgoing COR lists applied to the dial peer. II no
incoming COR list is configured, the inbound calls that are matched by the dia peer have
maximum permissions. If no outgoing COR list is configured, all outbound calls matched by
the dial peer are allowed.
J 2010 Cisco Systems. Inc
Verifying SIP Endpoint COR Settings
This subtopic explains how to view COR lists applied to SIP endpoints.
4-142
router# show voice register pool 1
Pool Tag 1
Config:
Mac address is 0024.C445.5233
Type is 7955
Number list 1 : DN 1
Proxy Ip address is 0.0.0.0
Class of Restriction List Tagi default
incoming corlist name is executive
DTMF Relay is disabled
Call Waiting is enabled
DnD is disabled
The show voice register pool command displays the parameters of all (using the all kevword)
or selected SIP endpoints. The parameters include the incoming and outgoing COR lists applied
to the endpoint. If no incoming COR list is configured, all calls that are originated by the
endpomt ha\ emaximum permissions. If no outgoing COR list is configured, all sources are
allowed toreach theSIPendpoint.
32010 Cisco Systems,
Summary
Summary
Calling privileges are used to prevent toll fraud and impose
other policy-defined dialing restrictions.
Voice gateways implement calling privileges byapplying
COR lists to inbound and outbound dial peers.
Cisco Unified CommunicationsManager Express and Cisco
Unified SRST routers implement COR by applying COR lists
to dial peers and SIP and SCCP endpoints.
Call is permitted whenthe incoming CORliston the inbound
dial peer or originating endpoint is a superset ofthe outgoing
CORlist on outbound dial peer or destination endpoint.
Verification of calling privileges involvesthe viewing of COR
names and lists and examining howthe COR lists are applied
to endpoints and dial peers.
4-144 Implementing Cisco VoiceCommunications and QoS (CVOICE) v8.0 2010 Cisco Systems. Inc.
Module Summary
This topic summarizes the key points that were discussed inthis module.
Module Summary
The numbering plandefines the basis for call routing and
should be made scalable to achieve robust operations.
The dial plan is the gateway call routing table and uses the
longest-match rule to findthe best path.
Digit manipulation adjusts calling and called numbers to the
requirements ofthe selected voice path (VoIP or POTS).
- Common path selection strategies include site-code dialing,
toll bypass, and TEHO.
Calling privileges prevent toll fraud and enforce policy-
defined dialing restrictions.
This module describes the concept of numbering and dial plans, explains how scalable plans are
designed, and liststhe benefits that scalability offersto the organization. It provides details
aboutcall routing principles that areemployed byCisco voice gateways, andpathselection
strategies whenintegrating VoIPand PSTN environments. The moduleshowsthat digit
manipulation playsanimportant roleindial planoperations andexplains digil manipulation
mechanisms available on voice gateways. It explains howto implement site-code dialing, toll
bypass, andtail-end hop-off(TEHO). anddescribes calling privileges thatare used to block
unwanted calls and to prevent toll fraud.
Module Self-Check
Usethe questions hereto reviewwhat you learnedin this module. The correcl answersand
solutions are found in the Module Self-Check Answer Key.
QI) Whichdial plancomponent is responsible for choosingtlieappropriate path for a call?
(Source: Introducing Call Routing)
A) endpoint addressing
B) call routing and path selection
C) call coverage and path selection
D) calling privileges
Q2) What is the dial plan component called endpoint addressing responsible for assigning
to the endpoints? (Source: Introducing Call Routing)
A) IP addresses
\i) E.164 addresses
C) gateways
D) directory numbers
Q3) Whichoptionimplements call routingandpathselectionon Cisco IOSgateways?
(Source: IntroducingCall Routing)
A) call routing tables
B) dialer maps
C) dial peers
D) route patterns
Q4) Which destination needs to have the PSTN access code 9 stripped from the called
number? (Source: Introducing Call Routing)
A) UCM
B) WAN
C) PSTN
D) gateway
Q5) What is one way to implement call coverage? (Source: Introducing Call Routing)
A) COR
B) pilot numbers
C) digit manipulation
D) endpoint addressing
Q6) Which three arc characteristics of a scalable dial plan? (Choose three.) (Source:
Introducing Call Routing)
A) backup paths
B) full digit manipulation
C) hierarchical numbering plan
D) dial plan logic distribution
E) granularity
F) high availability
Q7) What would be an appropriate dial-peer destination pattern for routing only national
calls in the United States? (Source: Understanding Dial Plans)
A) 9T
B) 9IXXXXXXXXXX
C) 91
I) I 9
08) Which three options are key requirements for a PSTN dial plan? (Choose three.)
(Source: Understanding Dial Plans)
A) internal call routing
B) inbound eall routing
C) outbound call routing
D) correct PSfN ANI presentation
F) internet call routing
F) digital \oice ports
Q9) What might some ISDN networks and PUXs expect along with a certain numbering
plan for both DNIS and ANI? (Source: Understanding Dial Plans)
A) foS
B) TON
C) OoS
D) CoS
QIO) Which command should be used to display infonnation for all voice dial peers?
(Source: Understanding Dial Plans)
A) show dial-peer voice summary
B) show dial-peer voice all
C) show dial-peer summary
D) show dial-peer all
Qll) Which function best deseribes a numbering plan? (Source: Understanding Dial Plans)
A) determines routes between source and destination
B) delines a telephone number of a voice endpoint or application
C) performs digit manipulation when sending calls to the PSTN
D) performs least-cost routing for VoIP calls
Q12) What are two t\pes of numbering plans? (Choose two.) (Source: Understanding Dial
Plans)
A) uni\ersal
B)
public
C)
private
D) U.S. public
Q13) Match the advantage ofa hierarchical numbering plan with itsdefinition. (Source:
Understanding Dial Plans)
A) simplified provisioning
B) simplified routing
C) summarization
D) scalability
F.) management
1. adds more high-level number groups
2. allows youto add newgroupsand modify existinggroupseasily
3. controls thenumber of groups from a single point intheoverall network
4. establishes a group of numbers in a specific geographical area or functions
group
5. keepslocal calls local and usesa specialized numberkey, suchas an area
code, for long-distance calls
QI4) Which worldwide prefix scheme was developed by the ITUto standardize numbering
plans? (Source: UnderstandingDial Plans)
A) E.164
B) G.1I4
C) G.164
D) E.I14
Q15) Whichentitiesare changed indigit manipulation? (Source: Describing Digit
Manipulation)
A) telephone numbers
B) IP addresses
C) voice gateways
D) dial peers
016) What happens by default when a gateway matches a dial string to an outbound POTS
dial peer? (Source: Describing Digit Manipulation)
A) The router strips off the left-justified digits that do nol explicitly match the
H) The router strips off the right-justified digits that explicitly match the
C) The router strips off the left-justified digits that explicitly match the destination
pattern.
D) The router strips off the right-justified digits that do not explicitly match the
Q17) By default. dial peers strip any outbound digits that explicitly match their
destination pattern. (Source: Describing Digit Manipulation)
A) PSTN
B) WAN
C) POTS
D) VoIP
>2010 Cisco Systems. Inc. Dial Plan Implementation 4-149
Q18) Gi\en the following dial-peer configuration, which two commands would ensure that
the complete numberis sent?(Choose two.) (Source: Describing Digit Manipulation)
destination pattern 1...
A) no digit-strip
B) forward-digits all
C) forward-digits 3
D) forward-digits 4
I ) no forward-digits
Q19) Gi\en the following dial-peer configuration, what would be the con'cct command to
add that would allow the call to complete to the PSTN? (Source: Describing Digit
Manipulation)
destination pattern 3...
port 0/0:23
A) prefix 5553000
B) prefix 5125552
C) prefix 95125552
D) prefix 5125553000
Q20) Which digit manipulation option is applied globally? (Source: Describing Digit
Manipulation)
A) number expansion
B| digit prefixing
C) digit forwarding
D) digit stripping
021) Which command is used to manipulate the ANI infonnation? (Source: Describing Digit
Manipulation)
A) did
B) clid-off
C) clid-on
D) ani
Q22) Which rule would search and replace a 10-digit numberwith the internal 2XXX
extension? (Source: Describing Digil Manipulation)
A) rule I /A2/ /4085552/
B) rule 1 /2//A4085552/
C) rule 1 /4085552//A2/
D) rulel/A4085552//2/
Q23) fhe command ofTers a simpler alternative to voice translation proliles when
expanding Cisco Unified Communications Manager Express extension numbers to
fuIK qualified public numbers. (Source: Describing Digit Manipulation)
Q24) The _ command displays the matching outgoing dial peer for a telephone number.
(Source: Describing Digil Manipulation)
Q25) Which option is used toaccomplish call routing and path selection? (Source:
Configuring Path Selection)
A) phone numbers
B) IP addresses
C) dia! peers
D) call managers
Q26) In Cisco IOS routers, which option is associated to each dial peer? (Source:
Configuring Path Selection)
A) call leg
B) translation rule
C) translation profile
D) interface
Q27) One best practice is to createa default POTS dial peer withthe direct-inward-dial
attribute usingthe ____ wildcardas the destination pattern. (Source: Configuring Path
Selection)
A)
B) #
C)
D)
Q28) Which principleenablesthe toll bypassfeature? (Source: Configuring PathSelection)
A) digit manipulationwhen sending PSTNcalls through IP WAN
B) backup dial peers for intersite calls
C) break-out into PSTN at the nearest point
D) dial peers for intersite calls
Q29) Which method is used to overcome the problem of overlapping directory' numbers?
(Source: Configuring Path Selection)
A) site code dialing
B) a technology prefix
C) TEHO
D) toll bypass
Q30) Instead of only routing intersite calls over an IP WAN link, also uses the IP
WAN link for PS'I'Ncalls. (Source: Configuring Path Selection)
Q31) What are two potential problems when configuring tail-end hop-off? (Choose two.)
(Source: Configuring Path Selection)
A) setting the DNIS
B) setting the ANI
C) redirecting the call through the WAN
D) sending the call into the PSTN on the breakout gateway
E) ensuring that the call arrives at the PSTN destination
F) ensuring proper callback operations
Q32) Which of these options is used within a dial plan lo define the destination that a user is
allowed to call? (Source: Configuring Calling Privileges)
A) dial peers
B) calling peers
C) calling privileges
D) destination patterns
Q33) When does a call fail? (Source: Configuring Calling Privileges)
A) COR list missing on the incoming dial peer
B) CORlist missing on the outgoing dial peer
C) COR list missing on either incoming or outgoing dial peer
D) CoS configuration missing on the incoming dial peer
034) Which COR component includes members that have been previously defined? (Source:
Configuring Calling Privileges)
A) dial peer
B) COR tag
C) dial tag
D) COR list
E) COR label
Q35) In Cisco I'nified Communications Manager Express, to which two entities are COR
lists assigned? (Choose two.) (Source: Configuring Calling Privileges)
A) ephone
B) ephone-dn
C) dial-peer
D) member
E) voice register pool
F) voice register director.' number
Q36) Which command is used to displav COR lists and members? (Source: Configuring
Calling Privileges)
A) show eor
B) show dial-peer cor
C) show dial-peer
D) show corlist
Q37) What is the objective of call coverage? (Source: Configuring Cal! Coverage)
A) Calls are answered as soon as possible.
B) Calls are distributed evenly belween group members.
C) Calls are not left unanswered.
D) Unanswered culls are sent to voice mail.
Q38) Which director)' number tvpe allows 12 concurrent calls to the same number? (Source:
Configuring Call Coverage)
A) dual-line director, number that is assigned to multiple ephones
B) octo-line director.' number that is assigned to multiple ephones
C) shared-line directory number that is assigned to multiple ephones
I)) o\ erlaid directory number that is assigned to multiple ephones
F.) all ofthe above
F) none ofthe above
4-152 Implementing Cisco Voice Communications and QoS (CVOICE| v8.0 2010 Cisco Systems. Inc
QD
B
Q2) D
Q3) C
04) c
Q5)
B
Q6) A, CD
Q7>
C
Q8) B,C, D
09) B
QIO) A
Qll) B
012) B.C
Q13) l-D
2-A
3-E
4-C
5-B
014] A
QI5) A
Q16) C
Q17) C
QI8) B, D
Q19) B
Q20) A
Q2D A
Q22) D
023} dialplan-pattern
024) show dialplan number
Q25) C
Q26) A
027) D
028) U
Q29) A
Q30| TEHO
Q31) B.F
Q32) C
Q33| D
034) D
Q35) B,K
Q36i B
)2010 Cisco Systems, Inc. Dial Plan Implementation 4-153
Q37) C
Q3S) D
4-154 Implementing Cisco Voice Communications and QoS(CVOICE) 8.0 2010 Cisco Systems, Inc
C_\K^^ sa
lable of Contents
Volume 3
Gatekeeper and Cisco Unified Border Element Implementation tl
5-1
Overview
Understanding Gatekeepers
Objectives 5-3
Gatekeeper Overview 5-4
Gatekeeper Functions 5-5
Gatekeeper Signaling 5~6
H.225 RAS Messages 5-7
Registration Request 5-10
Lightweight Registration 5-11
Admission Request 5"12
Location Request 5-14
H.225 RAS Intrazone Call Setup 5-16
H.225 RAS Interzone Call Setup 5-17
Gatekeeper Call Routing 5-18
Zones 5-18
Zone Prefixes 5-18
Zone Prefixes 5-19
Technology Prefixes 5-20
Technology Prefix Usage 5-21
Gatekeeper-Based Call Admission Control 5-22
Calculating Bandwidth 5-23
Configuring Gatekeeper 5-24
Gatekeeper Basics 5-25
Configuring Gatekeeper Zones 5-26
Configuring Remote Gatekeeper Zones 5-27
Gatekeeper ZonesConfiguration Example 5-28
Configuring ZonePrefixes 5-29
ZonePrefix Configuration Example 5-30
Configuring Technology Prefixes 5-31
Technology Prefix Configuration Example 5-32
Adapting H.323 Gateways toGatekeepers 5-33
Managing E.164 Address Registration 5-36
Gateway Configuration Example 5-37
Configuring Gatekeeper CAC 5"38
Gatekeeper CAC Configuration Example 5-39
Verifying Basic Gatekeeper Functionality 5-40
Verifying Gatekeeper Status 5-41
Verifying Registered Endpoints 5-42
Verifying Zone Prefixes 5-43
Verifying ZoneStatus 5-44
Verifying Gatekeeper Calls 5-45
Summary 5-46
Examining Cisco Unified Border Element 5-47
Objectives 5-47
Cisco Unified Border Element Overview 5-48
Cisco Unified Border Element Placement 5-49
Cisco Unified Border Element Applications 5-50
Cisco Unified Border Element Application Examples 5-53
Protocol Interworking on CiscoUnified BorderElement 5-54
Signaling Method Refresher 5-55
Cisco Unified Border Element Protocol Interworking 5-56
Media Flows on Cisco Unified Border Element 5-57
Media Flows 5-58
Cisco Unified Border Element Codec Filtering 5-59
Cisco Unified Border Element Codec Filtering Examples 5.50
Configuring Media Flow and Transparent Codec 5_61
Media Flow-Around andTransparent Codec Example 5-62
RSVP-Based CAC on Cisco Unified Border Element 5 63
RSVP-Based CAC 5.64
RSVP-Based CAC Call Flow 5.65
Cisco Unified Border Element Call Flows 5_66
SIPCarrierInterworking 5.67
SIPCarrierInterworking Call Flow 5.68
SIPCarrier Interworking with Gatekeeper-Based CAC Call Setup 5-69
Configuring H.323-to-H323 Interworking 5_70
Configuring H.323-to-H323 Interworking 5.71
Configuring H.323-to-H323 Fast-Start-to-Slow-Start Interworking 5.72
H.323-to-H323 Interworking Example 5_73
Configuring H.323-to-SIP Interworking 5_74
Configuring H.323-to-SIP DTMF RelayInterworking 5.75
Verifying Cisco Unified Border Element 5.76
Debugging Cisco Unified Border Element Operations 5-77
Viewing Cisco Unified Border Element Calls 5-78
Summary 5_79
Module Summary 5_81
Module Self-Check 5.33
Module Self-Check Answer Key 5-86
Quality of Service q-1
Overview 6-1
Introducing QoS ^3
Objectives 6-3
QoS Issues 6-4
After Converged Networks 6-5
Quality Issues in Converged Networks 6-6
Lack of Bandwidth 6-7
Managing Available Bandwidth 6-8
End-to-End Delay 6-9
Types of Delay 6-10
Reducing Delay 6-11
Packet Loss 6-12
Preventing Packet Loss 6-13
QoS and Voice Traffic 6-14
QoS Policy 6-15
QoS for Unified Communications Networks 6-16
Step 1: Identify Trafficand Its Requirements 6-17
Step 2: Divide Traffic into Classes 6-18
Step 3: Define Policies for Each Traffic Class 6-19
QoS Requirements 6-20
QoS Requirements: Video Telephony 6-21
QoS Requirements: Data 6-22
Methods for Implementing QoS Policy 6-23
Implementing QoS Traditionally Using CLI 6-24
Implementing QoS with MQC 6-25
Implementing QoS with Cisco AutoQoS 6-26
Comparing QoS Implementation Methods 6-27
QoS Models 6-28
Best-Effort Model 6-29
IntServ Model 6-30
DiffServ Model 6-31
QoS Model Evaluation 6-32
Summary 6-34
'i implementing CiscoVoice Communications and QoS (CVOICE) vS.O 2010CiscoSystems, Inc
w*
_m
Understanding QoS Mechanisms and Models . . 2222
^~~7^ 6-35
Objectives Rofi
DiffServ Model ~
DiffServ Model 1"^
DSCP Encoding "q
DiffServ PHBs
Expedited Forwarding PHB
Assured Forwarding PHB
DiffServ Class Selector
DiffServ QoSMechanisms
Classification .-
Marking fi.
Congestion Management
Congestion Avoidance J?
Policing ^^
Shaping
Compression
Link Fragmentation and Interleaving "^
Applying QoS to Input and Output Interfaces -54
Cisco QoS Baseline Model "55
Cisco Baseline Marking "^
Cisco Baseline Mechanisms ~57
Expansion and Reduction of Class Model "5
Summary 6"59
Explaining Classification. Marking, and LinkEfficiency Mechanisms ill
Objectives j*jj?
Modular QoS CLI
MQC Components
Configuring Classification 6-66
MQC Classification Options 6-67
Class Map Matching Options 6-69
Configunng Classification with MQC 6"70
Configuring Classification Using Input Interface and RTP Ports 6-72
Configuring Classification Using Marking 6"73
Configuring Class-Based Marking Jj-74
Class-Based Marking Overview 6-75
Configuring Class-Based Marking 6-76
Class-Based Marking Configuration Example 6'77
Trust Boundaries 6-78
Trust Boundary Marking 6-79
Configuring Trust Boundary 6-80
Trust Boundary Configuration Example 6-81
Mapping CoS to Network Layer QoS 6-82
_W Default LAN Switch Configuration 6"83
Mapping CoS and IP Precedence toDSCP 6"84
CoS-to-DSCP Mapping Example 6-85
DSCP-to-CoS Mapping Example 6-86
<_* Configuring Mapping 6-87
Mapping Example -88
Link Efficiency Mechanisms Overview 6-89
Link Speeds and QoS Implications 6-9
'__ml Serialization Issues 6-91
Serialization Delay 6"92
Link Fragmentation and Interleaving 6-93
Fragment Size Recommendation 6-94
Configuring MLP with Interleaving 6-95
Configuring MLP with Interleaving 6-96
MLP with Interleaving Example 6-97
Configuring FRF.12 Frame Relay Fragmentation 6-99
>2010 Cisco Systems. Inc Implementing Cisco Voice Communications andQoS(CVOICE) v8.0
6^0
6-41
6-44
Configuring FRF.12 Fragmentation 6-100
FRF.12 Configuration Example g.101
Class-Based RTP Header Compression 6.103
RTP Header Compression Example 6-104
Configuring Class-Based Header Compression 6-105
Class-Based RTP Header Compression Configuration Example 6-106
Summary g 107
Managing Congestion and Rate Limiting 6.109
Objectives 6-109
Congestion and Its Solutions 6-111
Congestion and Queuing: Aggregation 6-112
Queuing Components 6-113
Software Interfaces 6-115
Policing and Shaping 6-116
Policing and Shaping Comparison 6-118
Measuring Traffic Rates 6-119
Single Token Bucket 6-121
Class-Based Policing 6-123
Dual Token Bucket Single Rate Class-Based Policing 6-124
Dual Rate Class-Based Policing 6-125
Configuring Class-Based Policing 6-127
Configuring Class-Based Policing 6-128
Class-Based Policing Example: SingleRate, SingleToken Bucket 6-129
Class-Based Policing Example: Single Rate, Dual Token Bucket 6-130
Class-Based Shaping 6-131
Configuring Class-Based Shaping 6-132
Class-Based Shaping Example 6-133
Hierarchical Class-Based Shaping with CBWFQ Example 6-134
Low-LatencyQueuing 6-135
LLQ Architecture 6-136
LLQ Benefits 6-137
ConfiguringLLQ 6-138
LLQ Configuration Example 6-139
Monitoring LLQ 6-140
Calculating Bandwidth for LLQ 6-141
Summary 6-143
UnderstandingCisco AutoQoS 6-145
Objectives 6-145
Cisco AutoQoS VoIP 6-146
Cisco AutoQoS VoIP Functions 6-147
Cisco AutoQoS VoIP Router Platforms 6-148
Cisco AutoQoS VoIP Switch Platforms 6-149
Configuring Cisco AutoQoS VoIP 6-151
Configuring Cisco AutoQoS VoIP: Routers 6-153
Configuring Cisco AutoQoS VoIP: Switches 6-154
Monitoring Cisco AutoQoS VoIP 6-155
Monitoring AutoQoS VoIP: Switches 6-157
Automation with Cisco AutoQoS 6-158
Cisco AutoQoS for the Enterprise 6-159
Configuring Cisco AutoQoS for the Enterprise 6-161
Monitoring Cisco AutoQoS for the Enterprise 6-163
Monitoring Cisco AutoQoS for the Enterprise: Phase 2 6-164
Summary 6-165
Module Self-Check Answer Key 6-179
Implementing CiscoVoice Communications and QoS (CVOICE) vS.O 2010CiscoSystems, Inc.
Module 5
Gatekeeper and Cisco Unified
Border Element
Implementation
Overview
Gatekeepers play a major part inmedium and large H.323 VoIP network solutions. Gatekeepers
allow for dial-plan scalability and reduce the need tomanage global dial plans locally. This
moduledescribes the functions of a gatekeeper and explainshowto configure gatekeepers to
interoperate with gateways.
Also, this module gives anoverview oftheCisco Unified Border Element and describes how to
implement a Cisco Unified Border Element within an enterprise network. ACisco Unified
Border Flement has theability tointerconnect voice and VoIP networks, offering protocol
interworking. address hiding, and securityservices.
Module Objectives
Upon completing this module, you will beable toexplain what gatekeepers and Cisco Unified
Border Elements are. howtheywork, and what features theysupport. This abilityincludes
Describe Ciscogatekeeper functions and configuration
Explain Cisco Unified Border Element features andconfiguration
5-2 Implementing Cisco Voice Communications and QoS (CVOICE) u8.0 )2010 Cisco Systems, Inc
Lesson 1
Understanding Gatekeepers
Overview
Agatekeeper isan optional element in an H.323 environment. Ilisdeployed in larger H.323
VoIP network solutions, where scalability becomes an issue. Gatekeepers offer improved dial-
plan manageability by moving call routing logic tothe gatekeeper and reducing the dial plans
maintained by H.323 gateways. H.323 gatekeepers provide additional functions, such as Call
Admission Control (CAC), which prevents oversubscription of WAN bandwidth by VoIP calls.
Objectives
Upon completing this lesson, you will be able todescribe the functions and operation of
gatekeepers and explain how toimplement gatekeepers, including address resolution and CAC.
Describe the functionality of gatekeepers in an H.323environment
Describe the signalingbetween gatewaysand gatekeepers
Explain the gatekeeper call routing process, and therelated elements, such as gatekeeper
zone, zone prefixes, technology prefixes, andE.164 aliases
Describe how a gatekeeper supports CAC functions
List the steps necessary toconfigure a multizone gatekeeper for local and remote /onecall
routing
Describe howto configure local and remotezones on a gatekeeper
Explain howto configure gatekeeper zone prefixes
Describe how to configure gatekeeper technology prefixes
Explain how loadapt configuration of an H.323 gateway toregister with agatekeeper
Describe how to configure CAC functions on a gatekeeper
Explain how toverify- that H.323 endpoints are registered properly and calls are correctly
routed across a gatekeeper
Gatekeeper Overview
This topic describes 11.323 gatekeepers and their role in 11.323 signaling.
Gatekeeper Overview
Typical gatekeeper functions:
Agatekeeper is an H.323 entity on the network.
* Agatekeeper provides these services:
Address translation
Call Admission Control for H.323 terminals, gateways,
and multipoint control units
* Primary functions are admission control, zone management,
and E.164 address translation.
* Gatekeepers are logically separated from H.323 endpoints
such as terminals and gateways.
Gatekeepers are optional devices in a network
A gatekeeper is an 11.323 entity on the network thai provides services such as address
translation and network access control for H.323 terminals, gateways, and multipoint control
units. The primary functions of a gatekeeper areadmission control, zone management, and
F.164 address translation. Gatekeepers arelogically separated from H.323 endpoints and
optional devices in an 11.323 network environment.
Gatekeepers areoptional nodes that manage endpoints inanH.323 network. The endpoints
communicate with thegatekeeper using the Registration. Admission, andStatus (RAS)
protocol.
Note The ITU-T specifies that, althougha gatekeeper is an optionaldevice in H.323networks, ifa
network does include a gatekeeper, all H.323 endpoints should use it.
Implementing Cisco Voice Communications and QoS (CVOICE] vS.O
Gatekeeper Functions
This subtopic describes gatekeeper functions.
Gatekeeper Functions
Mandatory Description
Address
resoLtjon
Amission
control
Zone
management
TranslatesH.323IDs(such as flwyt@domain.oom) and E.164
numbers(standard telephonenumbers) toendpoint IPaddresses
Controlsendpointadmissionintothe H.323network
Provideszone managementforail registeredendpoints n the zone
Optionaj | Description
Cal
authorization
Cal
management
Bandwidth
management
Accesses restrictions for certain terminals or gateways or have time-
of-day poicies restrict access
Keepsstate ofactivecallinformation anduses Ittoindicate busy
endpoints or redirect calls
Rejectsadmissions whenthe required bandwidth is not available
Gatekeepers have mandatory and optional responsibilities. The mandatory tasks are as follows:
Addressresolution: Calls originating within anH.323 network may use analias toaddress
the destination terminal. Calls originating outside theH.323 network and received bya
gateway may use an E.164 telephone number toaddress the destination terminal. The
gatekeeper resolves the alias orthe E. 164 telephone number into the IP destination address
onthe H.323 network. Theresolution is doneusing a translation tablethat is updated with
registration messages.
CAC: The gatekeeper controls the admission oftheendpoints into ihe H.323 network.
Three RAS messages areused forthispurpose: Admission Request (ARQ), Admission
Confirmation (ACF). andAdmission Reject (ARJ). Asubfunction of CAC is thebandwidth
control that manages endpoint bandwidth requirements. When registering with a
gatekeeper, anendpoint will specify itspreferred coder-decoder (codec). During H.245
negotiation, adifferent codec may berequired. Codec negotiation isperformed using these
three messages: Bandwidth Request (BRQ), Bandwidth Confirmation (BCF), and
Bandwidth Reject (BRJ).
Zonemanagement: Agatekeeper is required toprovide address translation, admission
control, andbandwidth control for terminals, gateways, and multipoint control units that
are located within its zone of control.
An H.323 gatekeeper canprovidethese optionalfunctions:
Call authorization: Basedon policiessuchas time-of-day, the gatekeeper can restrict
access to certain endpoints or gateways.
Call management: With this option, thegatekeeper maintains active call information and
uses it to indicate busy endpoints or to redirect calls.
Bandwidth management: With this option, thegatekeeper canrejectadmission when the
required bandwidth is not available.
Gatekeeper and Cisco Unified Border Element Implementation 5-5
Gatekeeper Signaling
This topic explains the message types thai are involved in gatekeeper-based H.323 signaling.
5-6
Gatekeeper Signaling Overvi
Gateway
Gatekeeper
H22b r^S(UDP)^^^feR225RAS [UDP)
Gateway
Cisco gatekeepers use 11.323 RAS protocol as theprimary call-signaling method. RAS is a
subsetofthe H.225 signaling protocol and is basedon User Datagram Protocol (UDP).
Signaling messages between gateways are11.225 call control, setup, or signaling messages.
H.225 call control signaling is usedto set up connections between H.323 endpoints. fhe ITU
H.225 recommendation specifies the use andsupport of 0.931 signaling messages. If no
gatekeeper is present. H.225 messages areexchanged directly between theendpoints.
H.245 is negotiated after a call is signaled between thegateways. 11.245, a control signaling
protocol in the 11.323 architecture, allows theexchange of end-lo-end H.245 messages between
communicatingendpoints. The H.245 control messages are carried over H.245 control
channels. The H.245 control channel is thelogical channel 0 andis permanently open, unlike
themedia channels, fhe messages that arecarried include messages toexchange capabilities of
terminals and to open and close logical channels.
Alter a connection has beenset up via 11.225 signaling, the 11.245 call control protocol is used
to resolve thecall media type andestablish themedia flow. The H.245 call control protocol
also manages the call after it has been established.
As thecall is set up. other port assignments are dynamically negoliated:
Real-1 ime Transport Protocol (RTP) ports are negotiated fromthe lowest number, fhe
range is 16384 to 32768.
The H.245 ICP port is negotiated duringH.225 signaling for a standardH.323 connection.
H.225 RAS Messages
This subtopiclists the H.225 RASmessage types.
H.225 RAS Messages
Discovery Location:
Gatekeeper Request (GRQ) LocallonRequest (LRQ)
Galekewer CorJirmsbon (GCF > Locallon Conflmamyi (LCF)
Geleleeper Bs|t (GRJ)
Location Reject 1LRJ)
Registration
Admission:
RegiSraoofi Retjuetf (RRQ)
. Regisraoor Confirmation(RCF]
- Regratrano" R*rect (RRJ)
AOmusionRequeS [AR0)
' Admission CMiTlimatlontACF}
- AdmissionRe|ect (ARJ)
Un registration
UnregotrBbonRequest (URQ)
' urreg4trawnCQnfiimarti;UCF)
Disengage
DiserGJOeReqLiest(DRO)
. Untejitniliwi Reject (URJ)
Request in Progress:
Resource availability
Request in Progress (RIP)
Resoj'ceAvailar>*t|lr>tlicitr>r{RAI)
Rejoice Availably Connrmanon (RAG)
- IntoimaKmReqiJesHIRG)
Bandwidth1 InfoimatwnRequest Response (IRR)
BarOwtfltti Request IBRO) Information Request Acknottledarnenl(IACK)
. BanJwidni Confirmation (KF)
- BardwUHi Reject (BRJ)
- ln!orman Request Negative Acknowledgment
(INACK)
The figure shows common RAS signal messages, which are initiated by agateway and
gatekeeper. RAS message types include those listed here:
Gatekeeper discovery messages: Anendpoint multicasts a gatekeeper discover}' request.
Tlie Gatekeeper Request (GRQ) message requests that any gatekeeper receiving it respond
with a Gatekeeper Confirmation (GCF) message granting it permission to register. The
Gatekeeper Reject (GRJ) message isa rejection of this request, indicating thatthe
requesting endpointshouldseekanothergatekeeper.
GRQ: Message is sent by an endpointto a gatekeeper.
GCF: Reply from a gatekeeper to anendpoint indicating thetransport address ofthe
gatekeeper RAS channel.
GRJ: Reply from a gatekeeper to anendpoint rejecting the request from the
endpoint for registration. TheGRJ message usually occurs because of a gateway or
gatekeeper configurationerror.
Terminal and gateway registration messages: The Registration Request (RRQ) message
is a request toregister from a terminal toa gatekeeper. If thegatekeeper responds with a
Registration Confirmation (RCF) message, the terminal will usetheresponding gatekeeper
for future calls. If the gatekeeper responds with a Registration Reject(RRJ) message, the
terminal must seek another gatekeeper with which to register.
RRQ: Sentfrom anendpoint to a gatekeeper RAS channel address. Included inthis
message is the technology prefix, if configured.
RCF: Reply from the gatekeeper confirming endpointregistration.
RRJ: Reply from the gatekeeperrejectingendpoint registration.
Gatekeeper and Cisco Unified Border Element Implementation
Terminal and gateway unregistration messages: The Unregistration Request (URQ)
message requests that the association between a terminal and agatekeeper be broken. Note
that the URQ request is bidirectional, that is. a gatekeeper can request a terminal to
consider itself unregistered, and aterminal can inform agatekeeper that it is revoking a
previous registration.
I RQ: Sent from an endpoint or a gatekeeper to cancel registration
1nregistration Confirmation (I'CF): Sent from an endpoint oragatekeeper lo
confirm an unregistration
I'nregistration Reject (t!RJ): Indicates that anendpoint wasnot preregistered with
the gatekeeper
Resource availability messages: The Resource Availability Indication (RAI) message isa
notification from a gatewav toa gatekeeper of its current call capacitv (breach ll-series
protocol anddatarate for that protocol. Upon receiving an RAI message, thegatekeeper
responds with a Resource Availability Confirmation (RAC) message toacknowledge its
reception.
RAI: Used by gatewavs to informthe gatekeeper whether resources are available in
the gateway to take on additional calls
RAC: Notification from the galekeeper to the gateway acknowledging receiptofthe
RAI message
Bandnidth messages: An endpoint sends a Bandwidth Request (RRQ) to itsgatekeeper to
request an adjustment incall bandwidth. The gatekeeper eithergrants the requestwitha
BCFmessage or denies it with a RRJ message.
BRQ: Sent by the endpoint to the gatekeeper requesting an increase or decrease in
call bandwidth
BCF: Sent by the gatekeeper confirming acceptance ofthe BRQ
BRJ: Sent by the gatekeeper rejecting the RRQ
Location messages: Location messages are commonly usedbelween interzone gatekeepers
to get the IP addresses of different /one endpoints.
~ Location Request (LRQ): Sent by a gatekeeper to the directory galekeeper to
request the contact infonnation for one or more E.164 addresses. An LRQ is sent
directly to a gatekeeper if one is known, or it is multicast to the gatekeeper discovery
multicast address.
Location Confirmation (LCF): Sent by a responding gatekeeper and contains the
call signaling channel or RAS channel address (IP address) of itself or the requested
endpoint. It uses the requested endpoint address when Directed Kndpomt Call
Signaling is used.
Location Reject (LRJ): Sent bv gatekeepers that received an LRQfor a requested
endpoint that is not registered or that has unavailable resources.
Call admission messages: The ARQmessage requests that an endpoint be allowed access
to the packet-based network by the gatekeeper. The requestidentitiesthe terminating
endpoint and the bandwidthrequired. The gatekeeper cither grants the request with an ACF
message or denies it with an ARJ message.
ARQ: An attempt bv an endpoint to initiate a call.
Implementing Cisco Voice Communications and OoS (CVOICE) v8.0 2010 Cisco Systems, Inc
ACF: An authorization by the gatekeeper toadmit the call. This message contains
the IP address ofthe terminating gateway orgatekeeper and enables the originating
gateway toinitiate call control signaling procedures.
ARJ: Denies tlie request from the endpoint togain access tothe network for ihis
particular call if the endpoint isunknown orinadequate bandwidth is available.
Disengage messages: When acall isdisconnected, the endpoint sends a Disengage
Request (DRQ) to ihe gatekeeper. The gatekeeper confirms (disengage confirmation
[DCF]) orrejects (disengage rejection [DRJ]) the request. Ifsent from an endpoint toa
gatekeeper, the DRQ message informs the gatekeeper that an endpoint is being dropped. IT
sent from agatekeeper toan endpoint, the DRQ message forces acall tobe droppedsuch
arequest will not be refused. The DRQ message isnot sent directly between endpoints.
DRQ: Sent from the endpoint toagatekeeper when acall isdisconnected
DCF: Confirms the DRQthat was sent by the endpoint
DRJ: Rejects the DRQ thatwassent by thegatekeeper
Request inProgress (RIP) message: The gatekeeper sends outan RIP message toan
endpoint orgateway toprevent call failures due toRAS message timeouts during
gatekeeper call processing. Agateway receiving an RIP message knows tocontinue towait
for a gatekeeper response.
Status messages: Agatekeeper uses anInformation Request (IRQ) todetermine the status
ofan endpoint. In itsInformation Request Response (IRR), the endpoint indicates whether
it is onlineor offline. Thegatekeeper may also replythat it understands the IRQ
(Information Request Acknowledgment [1ACK]) or thatit docs notunderstand therequest
(InfonnationRequest Negative Acknowledgment [INACK]).
IRQ: Sent froma gatekeeper to an endpoint requesting status.
Information Confirm (ICF): Sent froman endpointto a gatekeeper to confirm the
status.
IRR: Sentfrom anendpoint to a gatekeeper in response to an IRQ. Thismessage is
alsosent from anendpoint toa gatekeeper if thegatekeeper requests periodic status
updates. Gateways use theIRR loinform thegatekeeper about theactive calls.
IACK: Usedby the gatekeeper to respondto IRRmessages.
INACK: Used by the galekeeper to respond to IRRmessages.
)2010Cisco Systems, Inc. Gatekeeper and CiscoUnified Border Element Implementation 5-9
Registration Request
This subtopic explains thegatekeeper registration.
Registration Request
Registration is the process bywhich gateways, terminals,
and multipoint control units join a zone and inform the
gatekeeper of their IP and alias addresses.
Registration occurs after the discovery process.
Gateway registers with either: Gatekeeper
-- H.323 ID .^^.
E.164 address
RRQ/S ^^\. RF
Gateway A
When registering with thegatekeeper, thegateway submits its H.323 ID(if configured), the
attached E.164 addresses, or both. Cisco Unified Communications Manager Express registers
bv default E.164addresses of all registered Skinny Client Control Protocol (SCCP) and Session
Initiation Protocol (SIP) endpoints. Cisco IOS gateways register by default Ihe E.164 addresses
of all analogendpoints that are attached lo Foreign Exchange Station(FXS) ports.
Examples of 11.323 ID and E.164 addresses are as follows:
H.323 ID: gatewav name a,domain.com
E.164 address: 4085551212
Lightweight Registration
This subtopic describes lightweight registration.
Lightweight Registration
H.323v1 gateway sent full registration every30 seconds
H.323v2 gateway startswith full registration with gatekeeper
Gatewaynegotiatestimersfor lightweight registration
- Gateway sends lightweight registration
Every negotiated timeout
Similar to keepalive
The gateway sends an
RRQ message with
keepalive - true before the
TTL limer expires
Before H.323 version 2, Cisco gateways reregistered with the gatekeeper every 30 seconds.
Each registration renewal used the same process as the initial registration, even though the
gateway was already registered with the gatekeeper. This behavior generated considerable
overhead at the gatekeeper. H.323v2 defines a lightweight registration procedure that still
requires the full registration process for initial registration, but uses an abbreviated renewal
procedure toupdate Ihegatekeeper andminimize overhead.
Lightweight registration requires each endpoint tospecify aTime to Live (TTL) value in its
RRQ message. When agatekeeper receives anRRQ message with aTTL value, it returns an
updated TTL timer value in an RCF message tothe endpoint. Shortly before the TTL timer
expires, the endpoint sends anRRQ message with ihe Keepalive field settoTRUE, which
refreshes the existing registration.
An H.323v2 endpoint isnot required toindicate aTTL initsRRQ. Ifthe endpoint does not
indicate a TTL. the gatekeeper assigns oneand sends it tothegateway intheRCF message. No
configuration changes arc permitted during a lightweight registration, soall fields are ignored
other tlian the endpoint identifier, gatekeeper identifier, tokens, and TTL. With H.323vl.
endpoints cannot process the TTL field in the RCF. The gatekeeper probes the endpoint with
IRQs for a predetermined grace period to learn if the endpoint is still alive.
Gatekeeper and Cisco Unified Border Elemeni Implementation 5-1'
Admission Request
Thissubtopic explains the messages that are involved inthe CAC.
[mission Request
fialpk..r J 8Q155W* Gateway A
Gateway A
^^v H-225CaliSe,uP(TCP) ,'
H.24SCBB Salop (TCP)
^DurtRTPÛ^stfoam Vffi-'
Galeway B
The figure shows an ARQ. Before the call isset up. Gateway Asends an ARQ request lothe
gatekeeper. The gatekeeper checks the status ofcalled parly and sends either an ACT message
or an ARJ message. Intheexample, thegatekeeper sends anACT message. The H.225 call
setupwill be directly between the two gateways.
Admission messages between endpoints and gatekeepers provide the basis for call admissions
and bandwidth control. Gatekeepers authorize access to11.323 networks by confirming or
rejecting an ARQ.
ARQ Message Failures
It may not be clear from the RAS ARJ message whythe message was rejected. Ilere are some
basic ARJ messages andthereasons why these messages occur:
calledPartyNotRejiistered: This message is returned because the called parly either was
neverregistered or has not renewed its registration wilha keepalive RRQ.
invalidPermission: Thecall violates some proprietary policy within thegatekeeper that is
tv picallv set by the administrator ofthe network or by the gatekeeper. For example, onlv
certain categories of endpoints may beallowed lo usegateway services.
requestDcnied: The gatekeeper performs zone bandwidth management, and the bandw idth
that is required for this call would exceed the bandwidth limit of the/one.
callerNntRegistered: The endpoint asking for permissionto be admitted to the call is not
registered withthe gatekeeper from whom it is askingpermission.
routeCallToGalekecper; Theregistered endpoint hasbeen sent a setupmessage from an
unregistered endpoint. andthegatekeeper wishes to route the call signaling channel.
invalidEndpointldentificr: The endpoint identifier in the ARQ is not the one that the
gatekeeper assigned to this endpoint in the preceding RCF.
5-12 Implementing Cisco Voice Communicationsand QoS (CVOICE) vS.O
resource!, navailable: This message indicates that the gatekeeper docs not have the
resources such as memorv or administrated capacity, to permit the call. Il could possibly
also be used in reference to the remote endpoint, meaning that the endpoint is unavailable.
However, another reason may be more appropriate, such as the call capacity has been
exceeded, which would return a callCapacilyExceeded message.
securitvDenial: This message refers to the Tokens orCryptoToken fields. For example,
failed authentication, lack ofauthorization (permission), failed integrity, or the received
crypto parameters are not acceptable or understood. This message might also be used when
the password or shared secret is invalid or not available, the endpoint is not allowed lo use
aservice, areplay was detected, an integrity violation was detected, the digital signature
was incorrect, or the certificate expired.
qosControlNotSupported: The endpoint specified atransport quality of service (QoS) of
gatekeepcrControlled in its ARQ, but the gatekeeper cannot or will not provide QoS for
this call.
incompleteAddress: This is used for what is referred to as "overlapped sending/' Ifthere
isinsufficient addressing information in the ARQ, the gatekeeper responds with this
message. This message indicates that the endpoint should send another ARQ when more
addressing infomialion is available.
routeCallToSCN: This message means that the endpoint istoredirect the call toa
specified telephone number on the Switched Circuit Network (SCN) or public switched
telephone network (PSTN). This isonly used ifthe ARQ was from an ingress gateway.
aliaseslnconsistent: The ARQdestinationlnfo contained multiple aliases that identify
different registered endpoints. This isdistinct from destinationlnfo containing one ormore
aliases identifying the same endpoint plus additional aliases that the gatekeeper cannot
resolve.
excccdsCallCapacity: This message was formerly callCapacityFxceeded. It signifies that
the destination endpoint does not have the capacity to accept the call. This is primarily
intended for use with gateways that are version 4orlater that report their call capacity to
the gatekeeper.
undefinedReason: "fhis message isused only ifnone ofIhe other reasons are appropriate.
2010 Cisco Systems, Inc. Gatekeeper and Cisco Unified Border Elemenl Implementation 5-13
Location Request
This subtopic explains how gatekeepers locate other gatekeepers.
Location Request
LRQ messages are commonly used between
interzone gatekeepers to obtain the IP of different
zone endpoints
LRQs forwarded using one of two methods;
Sequential
Based on priority and cost
Slower routing
* Less signaling
Blast
To all matching gatekeepers
Response selected based on
priority and cost
Faster routing
More signaling
Gateway
An H.323 IRQ message is sent bv agatekeeper to another galekeeper lo request aterminating
endpoint. Based on the information that iscontained in the IRQmessage, the second
gatekeeper determines the appropriate endpoint.
Note
The gatekeeper sendsout an RIP messagetoanendpoint or gateway toprevent call
failures due toRAS message timeouts during gatekeeper call processing Agateway
receiving an RIP message will continue towaitfora gatekeeper response.
For gatekeeper redundancv- and load-sharing features, you can configure multiple gatekeepers.
The LRQs aresent either sequentially or toall gatekeepers at thesame time (blast).
Sequential forwarding of LRQs isthe default forwarding mode. With sequential LRQ
forwarding, the originating gatekeeper will forward an LRQ to the first gatekeeper in the
matching list. The originating gatekeeper will then wait for a response before sending an LRQ
tothenext gatekeeper on the list. If theoriginating gatekeeper receives an LCF while it is
waiting, it will terminate the LRQ forwarding process.
Ifyou have multiple matching prefix zones, vou may want toconsider using sequential I RQ
forwarding rather than blast LRQ forwarding. With sequential forwarding, you can configure
which routes are primary', secondary, and tertiary.
The top figure illustrates the sequential LRQ process. Galekeeper Awill send an LRQ first lo
Ciatekeeper B. Ciatekeeper Bwill send areply as cither an LCF oran LRJ toGatekeeper A. If
Ciatekeeper Breturns an LCF to Gatekeeper A. the LRQ forwarding process will beterminated.
If Ciatekeeper Breturns anLRJ toGatekeeper A. then Ciatekeeper Awill send anLRQ to
Gatekeeper C. Ciatekeeper Cwill return either an LCF or LRJ toGalekeeper A. Then.
Ciatekeeper Awill either terminate the LRQ forwarding process orstart the LRQ process again
with Gateway D.
Note With sequential LRQs, there is afixed timer when LRQs are sent. Even if Gatekeeper Agets
anLRJ back immediately from Gatekeeper B, itwill wait a fixed amount of time before
sending the next LRQ to Gatekeeper Cand Gatekeeper D. You can speed up this process
by using the Irq Irj immediate-advance timer command,
The bottom figure illustrates the blast LRQ sending. The blast option allows Gatekeeper Ato
simultaneously send LRQs to Gatekeeper B, Gatekeeper C, and Gatekeeper D. This option
speeds up the location process but requires more signaling.
)2O10 Cisco Systems, Inc Gatekeeperand Cisco Unified Border Element Implementation 5-15
H.225 RAS Intrazone Call Setup
This subtopic describes the H.225 RAS signaling involved in an intrazone call setup.
H225;Q931
Call Setup
H245
Capabilities
Negolialion
PSTN/
"""""Privats
Votce
1 initiate Call
H225
RAS
H225
RAS
10 Ringback Tone
16 Media (RTP)
H 323 Gatekeeper
13 Capabilities ExcJiange .
14 Master/Slave Determination I
15 Open Logical Channel
RTPS|ffirl
ARQ = Admission Request
ACF = Admission Confirm
Intheexample shown in the figure, both endpoints have registered wilh thesame gatekeeper.
Call flow with a gatekeeper proceeds as follows:
1. Acall is initiated. At thispoint, both theoriginating gateway andtheterminating gateway
have locatedand registered with the gatekeeper.
2. fhe gatewav sends anARQ to thegatekeeper to initiate theprocedure. The gateway is
configured withthe domain or address ofthe gatekeeper.
3. The gatekeeper responds tothe ARQ with anAC!'. Inthe confirmation, the gatekeeper
provides the IP address ofthe tenninating gateway.
4. Theoriginating gatewav initiates a basiccall setupto the terminating gateway.
5. Before theterminating gatewav accepts thecall, it sends anARQ to thegatekeeper togain
permission.
6. The gatekeeper responds affirmatively using the ACF message.
7. Theterminating gatewav proceedswiththe call setupprocedure bysendingthe next H.225
messages to the originating gateway and then exchanging with it the H.245capabilities.
Duringthis procedure, if the gatekeeper responds lo either endpoint with an ARJ to the ARQ.
the endpoint that receives the rejection terminates the procedure.
5-16 Implementing Cisco Voice Communications and QoS (CVOICE) u8.0
H.225 RAS Interzone Call Setup
This subtopic describes the 11.225 RAS signaling involved in an interzone call setup.
H.225 RAS Interzone Call Setup
ARQ - Admission Request
ACF Admission Confirm
LCF Location Request
LCF Location Confirm
In the example shown inthe figure, the gateways belong todifferent zones and are registered
wilh different gatekeepers. Thecall setupprocedure involves these messages:
1. A call is initiated.
2. The originating gateway sends an ARQ toits gatekeeper (GKI) requesting permission to
proceed andasking forthesession parameters fortheterminating gateway.
3. GKI determines from its configuration that the terminating gatewayis associated with
GK2.GK1 sends an LRQ to GK2.
4. GK2determines the IPaddressofthe terminating gateway and sends it backin an LCF.
5. If GKI considers thecall acceptable for security andbandwidth reasons, it maps theLCF to
an ARQand sendsthe confirmation (ACF) to the originating gateway.
6. Theoriginating gateway initiates a call setup to theterminating gateway.
7. Theterminating gateway acknowledges thereceipt of thecall setupusing the Call
Proceeding message.
8. Before accepting the incoming call, theterminating gateway sends anARQ toGK2
requesting permission to accept the incoming call.
9. GK2 admits the call and responds with a confirmation (ACF).
10. Theterminating gateway rings thedestination endpoint andthenproceeds through the
H.225 call setup and H.245 control function procedures until the RTP sessions areinitiated.
Gatekeeper Call Routing
Zones
fhis topic describes how H.323 gatekeepers route calls.
Gatekeeper Zones
Zones:
H.323 endpoints are grouped into zones.
Each zone has one logical gatekeeper that manages all
the endpoints in the zone.
Zone prefixes:
Azone prefix is the part of the called number that
identifies the zone to which a call goes.
Zone prefixes are usually used to associate an area or
country code to a configured zone
'fhe two primary gatekeeper routing concepts arezones and/.one prefixes.
A/one is defined as the set of H.323 nodes that arccontrolled bya single logical gatekeeper.
Gatekeepers that coexist on a network ma.v beconfigured so that Ihey register endpoints from
different subnets. Ihere can onlv beoneactive gatekeeper perzone. These zones can overlay
subnets, and one gatekeeper can manage gatewav s in one or more of these subnets.
Lndpoinls attempt todiscover a gatekeeper andconsequently thezone of which they arc
members by using the RAS message protocol.
Zone Prefixes
A/.one prefixdetermines to which zone calls are sent. For a zone, which is controlled bya
gatekeeper, the /one prefixeshelp routethe call to the appropriate endpoint. fhe zone prefixes
are typically area codes.
For example. GKI is configured with the knowledge that /one prefix 2l2xxxxxxx (that is. anv
address beginning with212and followed by sevenarbitrary digits) is handled by gatekeeper
GK2. When CiKI is asked to admit a call lo destination address 212-555-1111, it knows to send
theI.RQtoGK2.
5-18 Implementing Cisco Voice Communications and OoS (CVOICE] v8 0
Zone Prefixes
This subtopic describes gatekeeper zone prefixes.
Zone Prefixes
- Identifies the destination zone for the call
Determines if a call is routed to a remote zone or managed
locally
Azone prefix isthe part ofthe called number that identifies the destination zone for acall.
Zone prefixes are usually used toassociate an area code toaconfigured zone.
The gatekeeper determines ifacall is routed to aremote zone orhandled locally. In the figure,
when a call toarea code 300 originates inZone A, it must beforwarded to the gatekeeper in
Zone B. Calls to area code 200 are handled locally.
Zone prefixes determine the zone towhich a call must beforwarded.
Technology Prefixes
fhis subtopic describes technology prefixes.
Technology Prefixes
* Technology prefix:
Optional feature toenable morecal! routing flexibility
Groups endpoints of the same type together
Usually identified by"#" sign, but can be any E.164 string
Technology prefix with hop-off
Calls will be routed to a specified zone, regardless ofthe
zone prefix in the address
Gateways can register using a technology prefix.
If no technology prefix is included in the dialed number, a
default technology prefix can be used.
Gatekeeper will only route a call to a gateway with a
matching technology prefix.
Gatekeeper routing isenhanced bv technology prefixes and technology prefixes with hopoff.
Technologv prefixes are used todenote different types or classes of gateways. Thegatewavs
are thenconfigured to register withtheir gatekeepers withtheseprefixes. Forexample, voice
gateways might registerwithtechnologv prefix ]#, 11.320 gateways with2#, voice-mail
gateways with 3#. and so on. More than one gateway may register into the same /one
(configured with the same /one prefix). When that happens, thegatekeeper makes a random
selection among gatewav s ofthe same type. Thecaller, who knows thetype of device that they
are tr> ing to reach, can nowprepend a technology prefix to the destination address to indicate
the tvpe of gatewav to use to get to the destination.
Technology Prefix with Hopoff
"fhe other gatewav -tvpe feature is the abilitv to force a hopoffto a particularzone. Normallv,
when anendpoint or gatewav makes a call ARQ lo itsgatekeeper, thegatekeeper resolves the
destination address by first looking for the technolog} prefix. When that is matched, the
remainingstring is compared against known /one prefixes. If the address resolves lo a remote
/one. the entire address, includingboth technology and zone prefixes, is sent to the remote
gatekeeper in an LRQ, fhat remotegatekeeper then usesthe technology prefixto decidewhich
of itsgatewavs tohopoff. Thezone prefix determines therouting to a /one. andthetechnology
prefix detcmiines thegatewav' inthat /one. fhis behavior canbeoverridden byassociating a
forced hopoffzone with a particular technology prefix. Thisforces thecall lo thespecified /one
regardless ofthe zone prefix in the address.
5-20 Implementing Cisco Voice Communicalions and OoS (CVOICE) v8 0
Technology Prefix Usage
This subtopic explains the use oftechnology prefixes.
Technology Prefix Usage
Distinguish between gateways that havespecific capabilities
within a given zone
- Common to differentiate between gateways that support
terminals, video endpoints, or telephony devices
Forexample, 1#for voice callsand 2# forvideocalls
Ifthe callers know the type ofdevice that they are trying toreach, they can include the
technology prefix in the destination address toindicate the type ofgateway touse toget tothe
destination. For example, if a caller knows that address 2005551111 belongs toa regular
telephone, the destination address of 1#2005551111 can be used, where 1# indicates that the
address should beresolved bya voice gateway. When thegatekeeper receives anARQ for a
call to 1#200555! 111, it contacts the gatekeeper serving zone prefix 200. Thegatekeeper
serving zone prefix 200 searches for a local gateway that registered the E. 164 address
2005551111. If there is no E.164 match, the gatekeeper selects a gateway that is registered in
that zone for the technology prefix 1#. The address ofthedestination gateway is offered tothe
originating gateway (using LCF and ACF). When the call tol#200555l 111 reaches the
terminating gateway, it strips offthetechnology prefix before matching theoutbound dial peer.
Cisco gatekeepers use technology prefixes toroute calls when there isnoE.164 addresses
registered (by agateway) that matches thecalled number. Without E. 164 addresses registered,
theCisco gatekeeper relies onthese two options tomake the call-routing decision:
Technology prefix matches option: Thegatekeeper uses thetechnology prefix that is
appended inthe callednumber to select the destination gateway or zone.
Default technology prefixesoption: Thegatekeeper assigns a default gateway for routing
unresolved call addresses.
Thegatekeeper uses a default technology prefix for routing all calls thatdonothave a
technology prefix or forgateways thatdonot have a technology prefix defined. The gatekeeper
matches the technology prefix todecide which of itsgateways to hop off. This option is useful
if the majority of calls hop offona particular type ofgateway, sothatcallers nolonger prepend
a technology prefix to the address.
Gatekeeper-Based Call Admission Control
This topic describes how the 11,323 gatekeepers implement CAC.
Gatekeeper-Based Call Admission
Control
Authorizes calls if the network can handle them
- Static configuration of available resources
Provides CAC to these devices:
Cisco Unified Communications Manager
H.323 endpoint
Gatekeeper
If loo many calls go through the
WAN. voice quality may degrade
for all calls.
In converged networks, a certain amount of bandwidth should be allocated to VoIP calls. Aller
theprovisioned bandwidth hasbeen fully utilized, subsequent callsshould be rejected to avoid
oversubscription of priority queues, which wouldcause quality degradation for all voicecalls.
Ihis functionknown as CACis essential to guarantee good voice quality ina multisite
deployment. Thegatekeeper maintains a record of all active callssothat it canmanage
bandwidth in a zone.
CAC' regulates voice qualitv bylimiting thenumber of calls that canbeactive on a particular
link at thesame time. CAC does not guarantee a particular level of audio quality on thelink,
but it doesallow youlo regulate theamount of bandwidth that isconsumed byactive callson
the link.
fhe Cisco IOS gatekeeper can provide CAC between these devices:
Cisco UnifiedCommunications Manager
Cisco UnifiedCommunications Manager Fxpress
H.323 endpoint
Thegatekeeper requires a sialicconfiguration ofthe available resources. The gatekeeper cannot
assign variable resources, as is the case with Resource Reservation Protocol (RSVP).
Calculating Bandwidth
This subtopic explains how the gatekeepers calculate the bandwidth that is required per call.
Calculating Bandwidth
Gatekeeper calculates call bandwidth as double codec rate
- Ignoring overhead
- For all codecs
Formula for zone bandwidth calculation on a gatekeeper
- (Number ofcalls) *(Codec bandwidth) *2
Example: Three G.711 calls:
3 ' 64 * 2 = 384 kb/S
G.711
G.729
kb/s on Gatekeeper
128 kb/s
16 kb/s
The gatekeeper is not able to calculate the exact amount ofbandwidth that is consumed by a
call. Bandwidth depends not only on the codec, but also sampling rate, header compression,
and additional overhead. The gatekeeper approximates the call bandwidth by doubling the
codec rate. This calculation is used for all codec types.
For example, three simultaneous G.711 calls consume, from agatekeeper perspective:
3*64kb/s*2=384kb/s
Configuring Gatekeeper
This topic describes how to configure an H.323 gatekeeper on aCisco IOS router.
Gatekeeper Configuration Oven
Gatekeeper:
* Configure local and remote zones.
* Configure zone prefixes.
Configure technology prefixes.
- Enable the gatekeeper.
Gateway:
Configure gateways to use H.323gatekeepers.
Configure dial peers to use H.225 RAS protocol.
Follow these steps to configure a Cisco IOS gatekeeper:
Step 1 Configure local and remotezones on the gatekeeper.
Step 2 Configure zone prefixes for all zones where calls should be routed.
Step 3 Configure technology prefixes toprovide more flexibility incall routing.
Step 4 Friable the gatekeeper function.
Follow these steps toadapt a Cisco IOS 11.323 gateway lo interwork with a gatekeeper:
Step 1 Configure H.323 gateway parameters that arc required toregister wilh a gatekeeper.
Some parameters are mandatory, others are optional.
Step 2 Configure the dial peers to use the RASprotocol.
Gatekeeper Basics
"fhis subtopic describes the basic commands that are required to implement aCisco IOS
gatekeeper.
Gatekeeper Basics
router(config)#
gatekeeper
Enters gatekeeper configuration mode
router(config-gk)#
no shutdown
Enables the gatekeeper
Should be done when configuration complete
- Some parameters cannotbe modified with active registrations or
calls
Does not make the gatekeeper operational if:
- No local zones are configured
- Local zones use HSRP address and interface is standby
Default: disabled
The gatekeeper command enters the gatekeeper configuration mode. This mode allows the
configuration of all othergatekeeper-related settings.
The gatekeeper feature isdisabled by default. The gatekeeper does not have tobe enabled
before tlie gatekeeper settings are configured. In fact, it isrecommended that the gatekeeper
configuration iscompleted before bringing up the gatekeeper because some characteristics may
bedifficult toalter while the gatekeeper is running, asthere may beactive registrations orcalls.
The no shutdown command enablesthe gatekeeper, but it does not makethe gatekeeper
operational. Thetwoexceptions to thisareas follows:
If nolocal zones areconfigured, a noshutdown command places thegatekeeper in
inactive mode, waiting for a local zone definition.
If local zones aredefined tousea HotStandby Router Protocol (HSRP) virtual address, and
the HSRP interface is instandby mode, thegatekeeper goes into HSRP standby mode. Only
whenthe HSRPinterface is activedoes the gatekeeper go intothe operational up mode,
Galekeeper and Cisco Unified Border Element Implementation
Configuring Gatekeeper Zones
This topic describes how toconfigure galekeeper /ones.
Configuring Local Gatekeeper Zone^
router(config-gk)#
zone local zone-name domain-name Iras-IP-address]
Defines local zone.
Identifiedby zone or gatekeeper name, and domain name
Onlyone ras-tP-addressargument can be defined for all local
zones
Local zones cannot use different RAS IP address
- When configured in tne first zone definition, it can be omitted for
all subsequent zones that automatically pickup this address.
A gatekeeper /one can be either local or remote.
The zone local command defines a local /one. Multiple local zones canbe defined. The
gatekeeper manages all configured local /ones. The zone local command defines orchanges
the IPaddress that is usedbv thegatekeeper.
OnI\ one ras-IP-oddress argument can bedefined for all local zones. You cannot configure
each /onetouse adifferent RAS IP address. Ifyou define this inthe first zone definition, you
can omit it for all subsequent zones, which automatically pick upthis address, if vou set it ina
subsequent zone local command, it changes the RAS address ofall previously configured local
zonesas well. After it is defined, you can changeit by reissuing anyzone local command with
a different ras-IP-address argument. If theras-IP-address argument is an HSRP virtual
address, it automatically puts the gatekeeper into HSRP mode. In this mode, the gatekeeper
assumes standbv or active status, depending onwhether the HSRP interface isonstandby or
active status.
Youcannot remov c a local zone if there areendpoints or gateways thatare registered init. To
remove the local /one. shut downthe gatekeeper first lo force unregistration.
Themaximum number olTocal /.ones that aredefined ina gatekeeper should nol exceed 100.
5-26 Implementing Cisco Voice Communications and QoS (CVOICE) w8.0
Configuring Remote Gatekeeper Zones
"fhis subtopic explains how to configure remote zones on agatekeeper.
Configuring Remote Gatekeeper Zones
router(config-gk)#
zone remote other-zone-name other-domain-name other-
gatekeeper-lp-addreas [port-number] [cost cost-value
[priority priority-value]] [foreign-domain]
Statically defines remote zone
- Identified byzone or gatekeeper name, and domain name
Optional for DNS-resolvable zones
- DNS is appropriate for H.323 ID-based calls, not E.164
- Gatekeeper resolves address automatically
When several remote zones are configured, they can be ranked:
- Bycost (0-100), default: 50
- Bypriority (0-100), default: 50
- Zone with lower cost or higher priority is preferred over others
Default port number: UDP1719
The zone remote command statically defines a remote gatekeeper ora remote zone if a
Domain Name System (DNS) isnot used. The DNS-bascd resolution can be used for H.323 ID-
based calls but not for E.164 calls. ForE.164 address routing, zoneprefixes areconfigured on
the gatekeeper and aremote zone must be defined before configuring azone prefix for that
zone.
When there areseveral remote zones configured, they can beranked by cost and priority value.
Azone with a lower-cost value and a higher priority value isgiven preference over others. A
zone with lower cost or higherpriority is preferred over others.
The remote gatekeeper iscontacted overthedefault UDP port 1719.
Gatekeeper and CiscoUnified BorderElementImplementation 5-27
Gatekeeper Zones Configuration Example
This sublopic prov ides alocal and remote zone configuration example.
Gatekeeper Zones Configuration
gatekeeper
zone local ZoneA cleco com 10.1 .1.10
zone local ZoneB cisco com
zone remote ZoneC clsco.com 10. 1.1.12
no shutdown
Zone A
The figure shows a network with two gatekeepers: GKI and GK2. OKI manages two local
zones: Zone A and Zone B. GK2 manages Zone C.
GKI settings are shown in the sample configuration. Ithas both local zones configured. The
first zone local command includes the gatekeeper local address. Zone Cisconfigured using the
zone remote command and points lo the IP address of GK2.
5-28 Implementing Cisco VoiceCommunicationsand QoS (CVOICE) vS.O
J2010Cisco Systems, Inc.
Configuring Zone Prefixes
This topic describes how toconfigure zone prefixes.
Configuring Zone Prefixes
router(config-gk)#
zone prefix zone-.name el64-prefix [blast | seq] [gw-
priority priority gw-alias] ^____
Adds a prefix to the gatekeeper zone list
- Identified byname of a localor remote zone or gatekeeper
(defined by using a zone local or zone remotecommand)
E.164 prefixcan include wildcards:
- Dot (.) matches a single character.
- Asterisk (*) matches any stnng.
* blast and seq options define the mode forsending LRQs
gw-priority defines preference for localzone gateways
- For calls to numbers beginningwith prefix e 164-prefix.
- Rangeis 0 (lowest, blocks the gateway) to 10 (highest); default
is 5.
- gw-altasname is the H.323 IDof a gateway.
Set with the h323-gateway voip h.323-id command.
Azone prefix isastring ofnumbers that isused to associate agateway toadialed number in a
zone. It is configured with the zone prefix command in gatekeeper configuration mode.
Agatekeeper can handle more than one zone prefix in one zone, but azone prefix cannol be
shared by more than one gatekeeper. Ifyou have defined a zone prefix asbeing handled by one
zone and then define it for another zone, the second assignment cancels the first.
The E.164-prefix parameter isastandard E. 164 prefix that isfollowed by dots (.). Each dot
represents adigit inthe E.164 address. For example, 200 ismatched by 20(1 and any seven
numbers. An asterisk(*) is a wildcardthat matchesanynumberof any digits.
The blast and scq keywords define the method ofsending LRQs if multiple hopoff gatekeepers
exist. The default is seq.
The gw-priority option defines how the gatekeeper selects gateways in its local zone for calls
tonumbers beginning with prefix el64-prefix. The range is from 0 to 10, where 0 prevents the
gatekeeper from using the gateway gw-alias for that prefix, and 10 places the highest priority
on gateway gw-alias. Tliedefault is 5.
The gu--alias name isthe 11.323 ID ofagateway that isregistered or will register with the
gatekeeper, 'fhisname issetonthe gateway with the h323-gateway voip h.323-id command.
When choosing a gateway, the gatekeeper first looks for ihelongest zone prefix match. Then it
uses the priority and the gateway statusto select fromthe gateways.
Zone Prefix Configuration Example
Ihis subtopic prov ides azone prefix configuration example.
Zone Prefix Configuration Exampl
gatekeeper
zone local ZoneA :isco com 10.1.1. If)
zone local ZoneB rlaco com
zone prefix ZoneA 2. . .
gw-prlorlty 5 GW-AI
zone prefix ZoneA 2. . . gw-prlorlty 10 GH -A2
zone prefix ZoneB 3. . .
no shutdown
Inthefigure, thegatekeeper with 1P address 10.1,1.10 manages twolocal zones: /one Aand
/one B. Zone Ahas the prefix 2,,. associated with it and /one Bhas prefix 3... asthe zone
prefix. Four digits areused bythegatekeeper for resolving theaddresses.
1he gateways GW-A1and GW-A2 are configured (nol shown in this figure) toregister into
/one A. The gatewav CiW-B isconfigured loregister into /one B. The gatekeeper has different
priorities that are assigned toGW-A1and GW-A2 that make GW-A1the preferred choice,
GW-A2 serves as the backup gatewav for Zone A.
When the gatekeeper receives anARQ for a call toa number inrange 3..., it will return the
address GW-B. if it successfully registered withthe gatekeeper.
Whenthegatekeeper receives an ARQfor a call to a numberin range2..., it will returnthe
address GW-A 1. if it registered with thegatekeeper. If GW-A I hasnot registered, and GW-A2
has registered, the gatekeeper returns the IP address GW-A2.
5-30 Implementing Cisco VoiceCommunicationsand QoS (CVOICE) vS0
Configuring Technology Prefixes
This topic describes how toconfigure technology prefixes.
Configuring Technology Prefixes
router(config-gk)fl
gw-type-prefix type-prefix [[hopoff gftidl] [hopoff gkid2]
[hopoff gkidn] [seq | blast]] [default-technology] [ [gw
ipaddr ipaddr [port] ] ]
Defines a technology prefix
- Recognized and stripped before checking forthe zone prefix
hopoff option specifies the hopoffgatekeeper:
Regardlessofthe zone prefix inthe destination address
- Multiple occurrences configure redundant gatekeepers
default-technology specifies that gateways registeringwith this
prefix are used as default for routing any addresses that are
otherwise unresolved.
gwipaddroption indicates thatthegateway is incapable of
registering technology prefixes. When thegateway registers, itis
added to the group for this type prefix.
To enable the gatekeeper toselect the appropriate hopoff gateway, use the gw-type-prefix
command toconfigure technology prefixes, which arealso known as gateway-type prefixes.
Select technology prefixes todenote different types or classes of gateways. Callers will need to
know the technology prefixes that aredefined andthetypeof device that theyare trying to
reach. Thisknowledge enables them toprepend the appropriate technology prefix to the
destination address forthetypeof gateway that is needed to reach thedestination.
Atechnology prefix is recognized and isstripped before checking forthezone prefix.
Technology prefixes should not lead toambiguity with zone prefixes. Such ambiguity can be
avoided byusing the# character to terminate technology prefixes, for example, 3#.
The hopoffgkid option specifies the gatekeeper where the call istohop off, regardless ofthe
zone prefix inthe destination address. The gkid argument refers toa gatekeeper previously
configured using the zone local orzone remote command. You can enter this keyword and
argument multiple times to configure redundant gatekeepers for a given technology prefix.
Thedefault-technology keyword is useful for declaring a specific gateway-type prefix as the
default gateway type tobeused for addresses thatcannot beresolved. Ifthemajority of calls
hop offona particular type of gateway, youcan configure the gatekeeper to use that technologv
prefix asthe default type sothat callers nolonger have toprepend a technology prefix onthe
address.
Thegwipaddroption isconfigured for a gateway thatis incapable of registering technology
prefixes. When it registers, it adds the gateway tothe group for this type of prefix, justasif it
hadsent thetechnology prefix in itsregistration. Thisparameter canberepeated to associate
more than one gateway with a technology prefix.
Technology Prefix Configuration Example
This subtopic prov ides a sample configuration fortechnology prefixes.
Technology Prefix Configurati
Example
Calls with no prefix
treated as with 1#
Calls to prefix 2# go to
Zone C without zone
prefix routing
192 168 1.1 does not
registertechnology
prefix
Zone A
GW-A1
gatekeeper
zone local ZoneA cisco com 10.1 1.10
zone local ZoneB cieco com
ZO ne local ZoneC cisco COm
tone prefix ZoneA 2... gw-prior ty 5 GW AI
zone prefix Zo neA 2 . . . gw-priority 10 GW-A2
zone prefix Zo neB 3...
3 type-prefix !# def suit-tech ology
gw type-prefix 2# hop aff ZoneC
gw type-prefix 99#* gw ipaddr 192.168 1 1
no shutdown
"î^^; Galekeeper ^^
m^mi 10.1 110 '5E"2*u
WAbL-JZ. GW.B 3x,.x
192 168.1.1
ZoneB
ZoneC
The figure illustrates how technology prefixes enhancecall routing:
fhe gw-t>pe-prefix 1#*default-technology command declares lit as the default
technologv. All calls without any technologv prefix configured will be treated as if 1# was
prepended in the called number.
fhe gw-tjpc-prefix 2#*hopoff ZoneCcommand routes all calls with thetechnology
prefix 2U toZoneC, Based on zone prefixes, such call redirection preempts call routing. In
fact. Zone C does not have a /one prefixconfigured at all.
Thegw -type-prefix 99#* gwipaddr 192.168.1.1 command declares thatthegatewav with
IP address 192.168.1.1 will have the technology prefix 99# associated with it when il
registers with thegatekeeper. This command is usedwhen thegateway cannot registerthe
technologv prefix itself, which may be the case on gatekeepers other than Cisco
gatekeepers.
Adapting H.323 Gateways to Gatekeepers
This topic describes how to adapt H.323 gateways lo interoperate with gatekeepers.
Adapting H.323 Gateways to Gatekeepers
rouCer(config-if) #
| h323-gateway voip bind arcaddr ip-address
Optional, sets the source IPaddressforoutgoing H.323 traffic
Affects H.225, H.245, and RAS messages
router (conf Ig-if )* ^^_^^__^_^_
h323-gateway voip interface
Mandatory, configures an interface as an H.323 gateway interface
Prerequisite forsettinggatewayID and referencing gatekeeper
Onlyone interface can be selected per gateway
router(conflg-lf1W
h323-gateway voip h323-id interface-id
Optional, sets the H.323 nameofthegateway that identifies this
gateway to its associated gatekeeper
Ifnot defined, gateway registers E.164 numbers, no H.323ID
Zl
The h323-gateway voip bind command isan optional command that binds the gateway feature
to an IPaddressof a networkinterface. It affects H.225, H.245, and RASsignaling. If the
command is notconfigured, therouter relies ontheIPlayer tochoose theoutgoing address.
Theh323-gateway voipinterface command configures aninterface asan H.323 gateway
interface. Thiscommand mustbeconfigured before setting thegateway IDandreferencing the
gatekeeper. Only oneinterface canbe selected pergateway.
TheH323-gateway voip h323-id command specifies the H.323 IDthatthe gateway will use
when registering with the gatekeeper. This command isoptional, and if not configured, ihe
gateway will register without providing its H.323 ID. This command does not affect whether
the gateway registers attached E.164addressesor not.
Adapting H.323 Gateways
Gatekeepers (Cont.)
router(config-if| #
h323-gateway voip id gatekeeper-id {ipaddr ip-addr [port-
number] multicast} [priority number]
Mandatory, defines the zone (or gatekeeper ID) lo registerwith
Gatekeeper ID must matchthe zone or gatekeeper ID configured on
gatekeeper
Case-sensitive
Multicast discovery listens onwell-known H323gatekeeperdiscovery
address 224.0.0 41, port UDP 1718
Prioritydefines order of alternate gatekeepers
router(config-if)#
h323-gateway voip tech-prefix prefix
Optional, defines the technology prefix that the gatewayregisterswilh the
gatekeeper
Affects routing of inbound calls
The h323-gateway voip idcommand specifies the gatekeeper toregister with, fhe galekeeper
IDdefined inthis command is configured on thegatekeeper using thezone local command. It
can bereferred toaseither galekeeper ID or /one ID. The command is mandatory for the
gatewav to attempt registration. The gateway can contact the gatekeeper usingcither its unicast
address or v\ell-known multicast address 224.0.0.41. The gatekeepers listen onUDP port 1718.
1hehJ23-gatenay voip tech-prcfi\ command defines the technology prefix that the gateway
registers withthegatekeeper. This settingaffectsthe gatekeeper call routingand can influence
the path selection for inbound calls.
Adapting H.323 Gateways to
Gatekeepers (Cont.)
router(config)#
gateway
=1
Mandatory, causes gateway todiscover and register, orunregisterwith
galekeeper
- GRQand RRQor URQ messages. Default: disabled
router(config-dlal-peer)#
seaaion target ras
Mandatory, points the H.323 session at gatekeeper
RAS signaling protocol
- Gatekeeper consulted totranslate E.164 address toIP address
Signaling senl toRAS UDP port 1719 instead ofH.225 UDP port 1720
Tlie gateway command, available in global configuration mode, causes the gateway to send out
gatekeeper discovery (GRQ) and registration (RRQ) requests. It must be configured on the
gateway to attempt registration. By default, the gateway feature isdisabled. The no gateway
command forces unregistration.
The session target ras command that isconfigured in dial peer mode directs the H.323
signaling atthe gatekeeper. This command instructs the gateway to use the H.225 RAS
protocol in addition to H.225.0-based signaling. RAS signaling uses UDP port 1719 instead of
UDP port 1720used for H.225.0.
5-35
Managing E.164 Address Registration
This subtopic describes how to manage L.164 addresses thai are registered by the gatewav with
the gatekeeper.
Managing E.164 Address Registrati
router(conflg-ephone-dn)#
number number [secondary number] [no-reg [both [ primary]
- no-reg keyword prevents E 164number registration ofephone-dns
Default: Both SCCP endpoints registered, SIPnot registered
register el64
Registers fully qualified E164numbers for POTS dial peerswith
FXS
Fully qualified 2001, not fully qualified:200....
Default. Enabled for active dial peers with FXS that is not shut down
dialplan-pattern tag pattern extension-length extension-
length [extension-pattern extension-pattern | no-reg]
* no-reg keyword prevents expanded number registration
Default Expanded numbersare registered at gatekeeper
Cisco Unified Communications Manager Fxpress registers, bydefault, the E. 164 numbers of
SCCP endpoints. The numbercommand, available inephone-dn configuration mode, has the
no-reg option that disables the E. 164 number registration with the gatekeeper. The suboption
both prev ents both the primary and sccondarv numbers from being registered. The suboption
primary prevents primary number registration.
Cisco voice gatewavs. bv default, register the fully qualified extension numbers that are
associated with analog endpoints attached toEXS ports. An example of a fully qualified
number is2001. while a not fully qualified number is20.... The register el64 command,
available indial peermode, controls theregistration of fully qualified extension numbers. It is
enabled by default. The no register e!64command prevents the registration of fully qualified
extension numbers of FXS-attached endpoints.
Cisco voice galewav s do not register numbers that are associated with any oilier types ofdial
peers, including SIP endpoints of Cisco Unified Communications Manager Express that are
represented as VoIP dial peers. This procedure is ditTerent from SCCP endpoints. which create
plain old telephone service (POTS) dial peers.
The dialplan-pattern command, available intelephony-service and voip register global
configuration mode, expands internal extension numbers topuhlic E. 164 addresses. By default.
Cisco Unified Communications Manager Express registers theexpanded numbers with the
gatekeeper. The command has the no-reg option, which disables the registration ofthe
expanded address.
5-36 Implementing Cisco VoiceCommunications and QoS |CVOICE| v8.0
Gateway Configuration Example
This subtopic provides asample configuration for agateway that registers with agatekeeper.
Gateway Configuration Example
interface Loopback 0
lp address 192.168.1.2 255.255.255.0
b.3 23-gateway voip interface
h323-gateway voip bind sreaddr 192.168.1.2
h323-gateway voip id ZoneA ipaddr 10.1.1.10
h323-gateway voip h323-id GW-A
h323-gateway voip tech-prefix 1#
destination pattern [3-7]...
session target ras
destination pattern 2101
no register 6164
port 0/1/0
I
gateway
1
number 2001 no-reg
2101
"WAT*
ZoneA
The figure illustrates the configuration ofagateway that uses the loopback 0interface as the
gateway interface. It sources signaling traffic from the Loopback 0IP address. The gateway
registers with the gatekeeper ID ZoneA that must exactly match the ID (case-sensitive)
configured on the gatekeeper using the zone local command. The gateway registers using the
gatewav H.323 ID ofGW-A and registers tech-prefix I#.
The calls to four-digit numbers starting with 3through 7will trigger an ARQ toward the
gatekeeper.
"fhe gateway has an analog endpoint that is attached to FXS port 0/1/0, but its extension
number will not beregistered asa result ofthe noregister el64 command.
The gateway is simultaneously aCisco Unified Communications Manager Express gateway. It
prevents the extension number 2001 ofthe ephone-dn 1from being registered with the
gatekeeper.
Gatekeeper and CiscoUnified BorderElement Implementation
5-37
Configuring Gatekeeper CAC
fhis topic describes how toconfigure gatekeeper-based CAC.
Configuring Gatekeeper C>
router(config-gk]#
bandwidth {check-destination | interzone | total |
session , remote} {default ! zone zone-name} bandwidth
1 Defines maximumaggregate bandwidth for H.323 traffic
interzone From a specific zone toail otherzones together
total1 Ail calls within one zone
session For a single session in a zone
remote. Toall remote zones together
default: Default value for each zone
Bandwidth-size defined in kb/s
check-destination; Destination zone bandwidth check before
responding to ARQ
Default only source zone and interzone values checked
default Unlimited maximumaggregate bandwidth
The CAC is implemented onthe gatekeeper using the bandwidth command. Il defines the
maximum aggregate bandwidth that isconsumed by 11.323-signaled calls, 'fhe scope isdefined
by these kevwords:
interzone: Iotal amount of bandwidth for H.323 trafficfrom the zone to all other /ones
total: Total amount of bandwidth for H.323 traffic thai is allowed within the zone
session: Maximum bandwidth that is allowed for a session in the zone
remote: Total amount of bandwidth lo all remote /ones
default: Default value fbr all zones: zone-specific value overwrites this setting
The bandwidth is specified in kilobits per second.
The destination-check kev word makes thegatekeeper check the bandwidlh available inthe
destination zone before responding tothe ARQ with an ACE. By default, the gatekeeper checks
only the bandwidth available in the source /one.
5-38 Implemenling CiscoVoice Communications and QoS (CVOICE) vS.O
2010Cisco Systems. Inc.
GatekeeperCAC Configuration Example
This subtopic provides asample CAC configuration.
Gatekeeper CAC Configuration
Example
All calls from Zone A to all other zones:
1024 kWs
Allcalls fromany other zone to all other
zones 512 kWs
All calls within Zone A: 2048 kb/s
Allcalls withinevery other zone. 1536
kbls
Ma* G 729 codec inZone A
Max G.711 codecin every other zone
Destination zone bandwidlh check
enabled
gatekeeper
zona local ZoneA clsco.com 10.1.1.10
zone local ZoneB cisco.com
zone local ZoneC cisco.com
zone prefix ZoneA 2...
zone prefix ZoneB 3...
zone prefix ZoneC 4...
bandwidth interzone ions ZoneA 1024
bandwidth interzone default 512
bandwidth total ions ZonaA 2048
bandwidth total default 1536
bandwidth session zone ZoneA 16
bandwidth session default 128
bandwidth chacle-destination
no shutdown
The figure illustrates how toconfigure agatekeeper tomeet these requirements:
All calls from Zone A to all other zones: 1024 kb/s
All callsfrom anyotherzone to all otherzones: 512kb/s
All calls within Zone A: 2048 kb/s
All calls within every other zone: 1536kb/s
Maximum G.729 codec in Zone A
Maximum G.711 codec in every other zone
Destination zone bandwidth check enabled
Gatekeeper and Cisco Unified BorderEfementImplementation 5-39
Verifying Basic Gatekeeper Functionality
This topic describes how toverily gatekeeper operations.
Gatekeeper Verification Overvi<
show commands.
show gatekeeper status
show gatekeeper endpoint
* show gatekeeper zone prefix
- show gatekeeper zone status
show gatekeeper calls
show gatekeeper gw-type-
prefix
debug commands:
debug h225 {asnl | eventsj
debug h245 {asnl | events}
* debug proxy h323 statistics
debug ras
debug gatekeeper main [5] [10]
The figure lists the commands that can be used tomonitor and debug galekeeper configurations
and interoperabilitywith gatewavs.
Verifying Gatekeeper Status
This subtopic explains how toverify' the gatekeeper status.
Verifying Gatekeeper Status
gk show gatekeeper status
Gatekeeper State: OP
Load Balancing! DISABLED
Plow Control: DISABLED
License Status: AVAILABLE
Zone Name: ZoneA
Zone Name: ZoneB
Accounting: DISABLED
Endpoint Throttling:
Security: DISABLED
Maximum Remote Bandwidth:
Current Remote Bandwidth:
unlimited
0 kbps
Current Remote Bandwidth
Hunt Scheme: Random
(w/ Alt GKs): 0 kbps
The show gatekeeper status command displays the operational and license status ofthe
gatekeeper. It lists all local zones, the maximum aggregate remote bandwidlh, and other options
that relate to functions, such as security and AAA.
Verifying Registered Endpoints
This subtopic illustrates how to verify ihe cndpoinls that are registered with the galekeeper.
Verifying Registered Endpoint!
gk# show gatekeeper endpoint
GATEKEEPER ENDPOINT REGISTRATION
CallSignalAddr Port RASSignalAddr Port Zone Name Type Flags
10.100.100.100 1720 10.100.100.100 56937 ZoneA
VOIP-GW
E1S4-ID: 2005551212
H323-ID: GW-A
Voice Capacity Max.= Avail.= Current.= 0
10.100.100.101 1720 10.100.100.101 49521 ZoneB
VOIP-GW
E164-ID: 3005551213
H323-ID: GW-B
Voice Capacity Max.= Avail.= Current.= 0
ITotal number of active registrations = 2
The show gatekeeper endpoint command displays the endpoints that are registered by each
11.323 gateway. The gatekeeper presents the IP address and /one name ofeach registered
gateway and lists the gateway H.323 ID(if available) wilh the fc.164 addresses attached to the
gateway on FXS ports or through SCCP endpoints.
5-42 Implementing Cisco VoiceCommunicalions and QoS (CVOICE) v8 0
Verifying Zone Prefixes
This subtopic explains how to verify the gatekeeper zone prefix configuration.
Verifying Zone Prefixes
gk# show gatekeeper zone prefix
ZONE PREFIX TABLE
Bit-NAME E164-PREFIX
ZoneA
ZoneB
ZoneC
ZoneD
ZoneD
200*
300*
400*
555
919*
The show gatekeeper zone prefix displays the zone prefixes configured with the zone prefix
commands. Ihe output effectively presents the gatekeeper call routing table that does not
include technology-prefix related mechanisms.
Gatekeeper andCisco Unified Border Efement Implementation 5-43
Verifying Zone Status
This subtopic explains how to verily the status ofgatekeeper zones.
Verifying Zone Status
gk# show gatekeeper lone status
GATEKEEPER ZONES
GK name Domain Name RAS Address PORT FLAGS
ZoneA cisco.com 10.1.250.102 1719 LS
QOS ATTRIBUTES :
DSCP Option : default
BANDWIDTH INFORMATION Ocbps) :
Maximum total bandwidth : unlimited
Current total bandwidth : 0.0
Maximum interzone bandwidth : unlimited
Current interzone bandwidth : 0.0
Maximum session bandwidth : unlimited
ZoneB
clsco.com 10.1.25 0.102 1719 LS
The show gatekeeper zone status command displays the status ofall zones that are configured
on the gatekeeper, including local and remote zones. The status includes general zone
information and additional parameters, such as bandwidth-relalcd settings.
5-44 Implementing CiscoVoice Communications and OoS (CVOICE) v8.D
Verifying Gatekeeper Calls
This subtopic describes how to verify the active calls that have been signaled using RAS
protocol.
Verifying Gatekeeper Calls
GK# show gatekeeper calls
Total number of active calls = 1.
GATEKEEPER CALL INFO
LocalCalllD
2-14476
Endpt(s): Alias
src EP: A-CDCME
Age(sees)
59
CallSignalAddr
192.168.3.254
Endpt(s): Alias
dst EP: ipipgw
CallSignalAddr
193.16B.1.3
Port
1720
Port
1720
BW
128 (kb/s)
E.164Addr
12005553001
RASSignalAddr
192.168.3.254
E.l64Addr
13005556666
RASSignalAddr
192.16B.1.3
Port
52668
Port
52060
The show gatekeeper calls command lists the active calls that have been signaled using RAS
protocol. The output includes information about the E.164 addresses involved in the call, the
bandwidth consumption (as computed by the gatekeeper), and the IP addresses ofthe gateways
that sienaled the call.
Gatekeeper and CiscoUnified BorderElement Implementation 5-45
Summary
his topic summarizes the key points that were discussed in this lesson.
Summary
* H.323 gatekeepers resolve addresses, provide Call
Admission Control, manage zones, and control bandwidth
utilized by endpoints.
Some RAS messages are exchanged between gateway and
gatekeeper (GRQ. GCF, GRJ, RRQ, RCF, RRJ, ARQ ACF
ARJ, URQ, UCF, URJ, DRQ, DCF, DRJ, RIP), while other
messages are exchanged between gatekeepers (LRQ LCF
LRJ).
1Gatekeepers route calls based onthe called number, which
may or may not contain a technology prefix.
Agatekeeper provides CAC byacceptingcalls that do not
exceed the maximum aggregate throughput.
Agatekeeper mustat least havea local zone configured to
become operational.
Summary (Cont.)
<Local zones define the zones served bythe local gatekeeper
while remotezones are zones controlled byother
gatekeepers.
Zone prefixes establish the call routing table ofa gatekeeper.
Technology prefixes affect gatekeeper call routing and can be
configured on gatekeepers and gateways.
H.323 gateways must be adapted to interoperability with
gatekeepers by commands in interface and dial peer
configuration mode.
Gatekeeper CAC is implemented using the bandwidth
command.
Gatekeepers allow the verification of configuredzones,
registered endpoints, and routed calls.
5-46 Implementing CiscoVoice Communications and QoS (CVOtCE) v80
Lesson 2
Examining Cisco Unified
Border Element
Overview
TheCisco Unified Border Element is like a traditional voice gateway because it connects two
voice legs together. While atraditional voice gateway interconnects aplain old telephone
service (POTS) call leg with a POTS orVoIP call leg, the Cisco Unified Border Element
interconnects two VoIP call legs. This lesson describes the concepts and features ofa Cisco
Unified Border Element inenterprise environments. It explains how toimplement the Cisco
Unified Border Element on a Cisco IOS router.
Objectives
Upon completing this lesson, you will be able to describe the functions and operation ofCisco
Unified Border Element, including address hiding, Call Admission Control (CAC), and
protocol and media interworking. This ability includes being able tomeet these objectives:
Describe the functionality ofa Cisco Unified Border Element and itsapplications in
enterprise VoIP environments
Explain how protocol interworking isperformed on aCisco Unified Border Element and
what interworkingoptions are supported
Describe howmedia flows are managedby a CiscoUnifiedBorder Element
Explain how Cisco Unified Border Element can be used toperform RSVP-bascd CAC
Describe cal! flowsin typical CiscoUnifiedBorder Element deployments
Explain how toconfigure H.323-to-H.323 interworking on aCisco Unified Border Element
Describe how toconfigure basic H.323-to-SIP interworking ona Cisco Unified Border
Element, including DTMF relay interworking
List the commands that are usedto configure media flow-around, media flow-through, and
transparent codec pass-through
Explain howto verify Cisco Unified Border Element operation
Cisco Unified Border Element Overview
This topic provides anoverview of Cisco Unified Border Element.
Cisco Unified Border Element Overvi
VoIP network interconnect
Also called session border controller
Ability to connect one VoIP dial peer with another VoIP
dial peer
Powerful protocol interworking toolset:
H.323-to-SIP
H.323-to-H.323
SIP-to-SIP
TheCisco Unified Border Element is anintelligent unified communications network border
element. ACisco Unified Border Element temiinatcs and reoriginates both signaling (H.323
and Session Initiation Protocol [SIP]) and media streams (Real-Time Transport Protocol [RTP]
and Real-) ime Transport Control Protocol [RTCP|) while performing border interconnection
services between IPnetworks. The Cisco Unified Border Element was formerly known asthe
CiscoMukisen ice IP-to-IP Gateuay. The CiscoUnified Border Element, in addition to other
Cisco IOS Software features, includes Session Border Controller functions that help enable
end-to-end. IP-based transport of \oiee. video, and data between independent Cisco Unified
Communications networks.
Originally. Session Border Controllers were used byservice providers toenable complete
billing capabilities within VoIP networks. However, the functionality tointerconnect VoIP
networks is becoming more andmore important for enterprise VoIP networks, because VoIP is
becoming the newstandard for anv telephony solution.
VoIP dial peers can bemanaged by cither SIP or 11.323. As a result, the ability to interconnect
VoIP dial peers includes the abilitv tointerconnect VoIP networks using different signaling
protocols, or VoIP networks using the same signaling protocols bul facing inleroperahility
issues.
Protocol interworking includes these combinations:
II.323-to-SIP interworking
Il.323-to-H.323 interworking
SIP-to-SIP interworkine
Cisco Unified Border Element Placement
This subtopic describes where Cisco Unified Border Element is typically placed within
enterprise networks. ^^^^
Cisco Unified Border Element Placement
SIP or H.323
connects\folP dial peers.
Inbound
\folP Dial Peer
Outbound
VoIP Dial Peer
Cisco Unified
Border Element
SIP or H.323
The figure illustrates the capability ofCisco Unified Border Element to interconnect VoIP
networks, including VoIP networks that use different signaling protocols. VoIP interworking is
achieved by connecting an inbound VoIP dial peer with an outbound VoIP dial peer.
The Cisco Unified Border Element provides anetwork-to-network interface point for the
following functions:
Signaling interworking (H.323 andSIP)
Media interworking (flow-through, flow-around, and dual tone multifrequency [DTMF])
Address and port translations (privacy and topology hiding)
Billing and Call Detail Record (CDR) normalization
Quality ofservice (QoS) and bandwidth management (QoS marking using differentiated
services code point [DSCP] ortype ofservice [ToS], bandwidth enforcement using
Resource Reservation Protocol [RSVP], and codec filtering)
ACisco Unified Border Element interoperates with many different network elements, including
voice gatewavs. IP phones, Cisco Unified Communications Manager, Cisco Unified
Communications Manager Express, and simpler toll bypass and VoIP transport applications.
The Cisco Unified Border Element provides organizations with all theborder controller
functions integrated into the network layer tointerconnect Cisco Unified Communications
voice andvideo enterprise-to-service provider architectures.
Gatekeeper and CiscoUnified Border Element Implementation 5-49
Cisco Unified Border Element Applications
This subtopic describes the tvpical applications of Cisco Unified Border Element in enterprise
environments.
Cisco Unified Border Element Applications
External connections.
Interconnect with VoIP carriers
Interconnect with other voice and video networks
Integrate Internet VoIP and video-over-IP users
Internal connections:
Increase interoperabilitywithin a VoIP network
Relevant features:
Protocol interworking
Address hiding
Security
video integration
CAC
Cisco Unified Border Elements serve these two main purposes in enterprise deployments:
External connections: ACisco Unified Border Elenicnl can be usedas a demarcation
point within aunified communications network and provides interconnectivity with
external networks. This purpose includes H.323 voiceand videoconnections and SIP VoIP
connections.
Internal connections: When usedwithin a VoIPnetwork, a Cisco Unified Border Element
can beused toincrease the flexibility and interoperability between different devices.
Following are somekev features that are offered by CiscoUnified Border Element;
Protocol inleniorking: The Cisco Unified Border Element supports interworking of
signaling protocols, including II.323-to-II.323. H.323-to-SII\ andSIP-to-SIP.
Address hiding: ACisco Unified Border Element can hide orreplace the endpoint IP
addresses used for the media connection.
Security: ACisco Unified Border Element can be placed inademilitarized /one(DMZ)
and prov ide outside connectiviiv to external networks.
Video integration: In addition to VoIPservices, a Cisco Unified Border Element also
supports H.323 video connections.
CAC: A Cisco Unified Border Element can use Cisco IOS Software-based CAC
mechanisms, including RSVP.
5-50 Implementing CiscoVoice Communications and QoS (CVOICE] v80
'2010 Cisco Systems, Inc
The table provides details ofthe key features.
Key Features of the Cisco Unified Border Element Gateway
Feature
Protocols
Network hiding
CAC
Protocol and signal
interworking
Media support
Media modes
Video codecs
Transport mode
DTMF
Fax support
Modem support
Details
H.323 and SIP
IPnetworkprivacy and topologyhiding
IP network security boundary
Intelligent IP address translation for call media and
signaling
Back-to-back useragent, replacing all SIP-embedded IP
addressing
RSVP
Maximum number of calls per trunk (maxcalls)
CAC based on IP circuits
CAC based on total calls, CPU usage, or memory usage
thresholds
H.323 toH.323 (including Cisco Unified Communications
Manager)
H.323 toSIP (including Cisco Unified Communications
Manager)
SIPtoSIP(including Cisco Unified Communications
Manager)
RTP and RTCP
Media flow-through
Media flow-around
H.261, H.263,andH.264
TCP
User DatagramProtocol(UDP)
TCP-to-UDP interworking
H.245 alphanumeric
H.245 signal
RFC 2833
SIP Notify
Keypad Markup Language (KPML)
Interworking capabilities:
H.323 to SIP
RFC 2833 to G.711 Inband DTMF
T.38 Fax Relay
Fax pass-through
Cisco Fax Relay
Modem pass-through
Cisco Modem Relay
Feature
Supplementary services
Details
Callhold, Call Transfer, and call forward for H.323
networks using H.450 and transparentpassingof Empty
Capability Set (ECS)
SIP-to-SIP supplementary services (holds and transfers)
support using the SIP REFER method.
H.323-tr>SIP supplementaryservices for Cisco Unified
Communications Manager with Media Termination Point
(MTP) on the H 323 trunk
Network Address
Translation (NAT)
Traversal
QcS
Voice-quality statistics
Number translation
Codecs
Transcoding
Security
Authentication.
authorization, and
accounting (AAA)
Voice media applications
Silling
NAT Traversal support for SIP phones that aredeployed
behind nonapphcation layer gateway (ALG) data routers
Stateful NAT Traversal
IP precedence and DSCP marking
Packet loss, jitter, and round-trip time
Number translation rules for VoIP numbers
Electronic Numbering (ENUM) support forE.164 number
mapping into Domain Name System(DNS)
G711 mu-law and a-law
G.723ar53, G.723ar63, G.723r53, and G 723r63
G 726M6, G.726r24, and G.726r32
G728
G.729, G.729A. G 729B, and G.729AB
Internet Low BitrateCodec (iLBC)
Transcoding between any two different families of codecs
from the following list:
G 711 a-law and mu-law
G.729, G 729A, G.729B. and G.729AB
G.723 (5.3 and 6 3 kb/s)
iLBC
IP Security (IPsec)
Secure RTP (SRTP)
Transport Layer Security (TLS)
AAA with RADIUS
Tool Command Language (Tel) scripts support for
application customization
VoiceExtensibleMarkup Language (VoiceXML, or
VXML 2 0) script support for application customization
Standard CDRs for accurate billing available through the
following
AAA records
Syslog
Simple Network Management Protocol (SNMP)
5-52 Implementing Cisco Voice Communicalions and QoS(CVOICE| v8.0
Cisco Unified Border Element Application Examples
This subtopic describes typical uses of Cisco Unified Border Element in an enterprise network.
The figure shows the various deployment options for aCisco Unified Border Element,
including an internal and an external connection. The internal connection provides connectivity
services between two sites ofthe same organization. Internal connections may utilize two Cisco
Unified Border Elements, which can be collocated with Cisco Unified Communications
Manager Express or aregular voice gateway. Two Cisco Unified Border Elements might be
needed if. for example, the two sites have acombination ofSkinny Client Control Protocol
(SCCP) and SIP phones, and H.323 is used over the WAN network. The external connection is
used to provide connectivity to the external Internet telephony service provider (ITSP).
Gatekeeper and CiscoUnified Border Element Implementation
Protocol Interworking on Cisco Unified Border
Element
Tîstopie describes the Cisco Unified Border Element protocol interworking capabilities.
Protocol Interworking on Cisco Unlfh
Border Element
Solves interoperability issues when using different signaling
protocol orwhen deviceshave different capabilities
Translates between signaling protocols:
Each call leg terminates on the Cisco Unified Border
Element.
The Cisco Unified BorderElement examines received
information, performs translation, and reoriginates a new
call leg.
Using interworking signaling protocols on Cisco Unified Border Element is like using aproxy.
1his feature can be used for two scenarios:
Interworking between the samesignaling protocols: ACisco Unified Border Element
that is using interworking between the same signaling protocols (for example H.323-to-
H.323) can be used to soke interoperability issues between two devices having different
capabilities. Because the Cisco Unified Border Element builds two different call legs to
each peer, it can work between those two call legs.
Interworking betweendifferent signaling protocols: ACisco Unified Border Element
can interconnect dial peers that use different signaling protocols, such asa SIP and an
H.323 dial peer. This allows for greater flexibility when deploying an IP communications
network.
5-54 Implementing Cisco Voice Communications and QoS(CVOICE| vS.O
2010 Cisco Syslems. Inc.
Signaling Method Refresher
fhis subtopic provides areview of signaling methods in H.323 and SIP.
Signaling Method Refresher
Stow start
Fast start
(Cisco default)
Eartymedia H.323
Delayedoffer SIP
H.323 wl
H.323 v2
Earlyoffer SIP
(Cisco default)
Early media SIP
Characteristics
H.245 parameters exchanged after H.225 connect.
H.245 parameters exchanged earlier, in H.225 cal setup
and H.225 cal proceeding/alerting.
Earty media cut-through after H.245 exchanged.
SDP proposals sent late:
Fromterminating gateway:200 OK
From originating gateway ACK
SDP proposals sent earty:
Fromoriginafinggateway. Invite
- From terminating gateway;
-200 OK
-183 Session Progress, or
180 Ringing
Earlymedia cut-lhrough after:
-183 session progress, or
-180 ringing
The table in the figure provides areview ofthe signaling methods that are supported by H.323
and SIP.
H.323 version 1supports only slow-start call setup, inwhich the H.245 parameters were
exchanged after thecall has been answered.
H.323 version 2 introduced the fast-start option, used by default onCisco gateways, which
expedites the call setup by embedding H.245 parameters in H.225 Call Setup and Proceeding or
Alerting messages.
Early media is an H.323v2 capability that allows the endpoints to establish RTP media flows
before the call isanswered. This option requires that fast start isused, but fast start docs not
necessarily entail early media cut-through, because itisnegotiated separately.
Delayed offer is aSIP signaling method that exchanges Session Description Protocol (SDP)
information about the media types, codecs, and RTP numbers late inthe exchange, namely in
the 200 OK and ACK messages.
Early offer, which is used by default on Cisco gateways, expedites the call setup by attaching
the SDP information toearlier messages: Invite, and 200 OK, 183 Session Progress, or 180
Ringing. The relevant difference is that the INVITE message carries the SDP information
rather than the 200 OK message in delayed offer.
Early media in SIP is the conceptual equivalent ofearly media in H.323 and allows an earlier
cut-through ofRTP flows. Itrequires early offer but is not enforced by it, because itis
negotiated separately.
Cisco Unified Border Element Protocol Interworking
This subtopic explains the protocol interworking options that are supported bv Cisco Unified
Border Element.
Cisco Unified Bordei
Ini
H.323 H.323;
Slow Start ; -;
Slow Start
Fast Start - -
Fast Start
*"""" """*g^*-
Delayed Offer ;-_ ::.
Early Offer - """
Delayed Offer
Early Offer
* *- __ . -s- .,,...... i.
_^,r~%
^^
Slow Start "^-_J*
Fast Start '-?r**S*
Delayed Offer
Early Offer
" " *&g~
- -*-
SIP
SIP
H.323 SIP
y"
When you use interworking signaling protocols, aCisco Unified Horder Element supports these
combinations:
ll.323-to-II.323: all combinations offast start and slow start onboth call legs
l.323-to-SIP: H.323 fast start-to-SlP carh oflcr and H.323 slowstart-to-a SIPdelaved
offer
SIP-to-H.323: SIP carh' offer toH.323 fast start. SIP early offer toH.323 slow start, and
SIP delayed offer to 11,323slow start.
SIP-to-SIP: All combinations ofearly offer and delayed offer on both call legs
5-56 Implementing Cisco Voice Communications and QoS(CVOICE) v8.0
Media Flows on Cisco Unified Border Element
This topic describes how Cisco Unified Border Element manages media flows.
Cisco Unified Border Element Sigs
and Media Flows
Cisco Unified Border Element canact as a proxy for H.323
and SIP (proxy signaling).
Media flow-through (default)All media streams are routed
through theCisco Unified Border Element:
- Solves IP interworking issues
- Hides IP original addresses
- Enables tighter security policies
Media flow-around: Media streams flow directly between
endpoints.
- Supported only for H.323-to-H.323 and SIP-to-SIP
Because aCisco Unified Border Element is asignaling proxy, italso processes all signaling
messages that negotiate RTP media channels. This processing enables aCisco Unified Border
Element to affect the flow ofmedia traffic. Two options exist: media flow-through and media
flow-around.
When using media flow-through, aCisco Unified Border Element replaces the source IP
address that is used for media connections with its own IP addresses. This operation can be
utilized in the following ways:
It solves IP interworking issues, because the Cisco Unified Border Element replaces
potential duplicate IP addresses with asingle, easy-to-control IP address.
It hides the original endpoint IP address from the remote endpoints.
This makes aCisco Unitied Border Element with media flow-through ideal fbr interworking
with external VoIP networks and enforcing a tighter security policy.
When using aCisco Unified Border Element internally, media flow-through might nol be
necessary or even desirable. One ofthe main drawbacks when using media flow-through is the
higher load on aCisco Unified Border Element router, which decreases the number ol
supported concurrent flows. In addition, media flow-through might result in suboptimal traffic
flows, because direct endpoint-to-endpoint communication is prohibited. Thus aCisco Unified
Border Element can also beconfigured fbrmedia flow-around.
When using media flow-around, aCisco Unified Border Element leaves the IP addresses used
for the media connections untouched. Call signaling will still be processed by the Cisco Unified
Border Element, but after the call isset up, the Cisco Unified Border Element isno longer
involved with the traffic flow. Media flow-around is supported only by interworking within the
same signaling protocol (H.323 or SIP), and is not available for H.323-to-SIP interworking.
Media Flows
This subtopic describes Cisco Unified Border Element media (low options.
Media Flow-Through:
101 M | 10 111<> 192 163 12| 192 168 12 | 102 1B8.1 2<> 10.2 11| 102
Media Flow-Around:
Cisco Unified
Border Element
192 168 1 2
Signaling Signaling
101 1 1 <> 10.2 1 1
Cisco Unded
Communications
Manager Express
The figure compares Cisco Unified Border Element media flow-through with flow-around:
In media flow-through, the signaling between two Cisco Unified Communications Manager
clusters isprocessed b> the Cisco Unified Border Element. The source IPaddresses ofthe
endpoints are replaced b\ the Cisco Unified Border Element IP address. Both endpoints
have the same IPaddress, but due lotheproxy function ofthe Cisco Unified Border
Element, no interworking issuesarise.
In media flow-around, the endpoint IP addresses cannot overlap and IP address hiding is
not required. The Cisco Unified Border Element still processes all signaling traffic, but the
endpoints have direct media channels that areestablished between them.
5-58 Implementing Cisco Voice Communications andQoS(CVOICE) v8.0
Cisco Unified Border Element Codec Filtering
This subtopic describes the codec handling options that are supported by Cisco Unified Border
Element.
Cisco Unified Border Element Codec
Filtering
VolPnetworks support multiple codecs:
- Preferencesdefinewhich codecs are selected over
others.
Cisco Unified Border Element can limit codecnegotiation toa
single codec:
- Ensures that a specific codec is negotiated
- Simplifies designconsiderations
Cisco Unified Border Element transparent codec handling:
- Transparently passes codec capabilities between
endpoints
Implemented via dial peer configuration
VoIP networks usually support alarge variety ofcodecs, and mechanisms exist to perform
codec negotiations between devices. Regardless ofwhich mechanisms are used, preferences
determine which codecs will be selected over others.
Because a Cisco Unified Border Element is essentially aCisco IOS gateway with the ability to
interconnect VoIP dial peers, the same codec selection mechanisms arc available as onany
other Cisco IOS gateway. Adial peer can be configured toallow aspecific codec ortouse a
codec voice class to specify multiple codecs with apreference order. This configuration enables
aCisco Unified Border Element toperform codec filtering, because adial peer will only setup
acall leg ifthe desired codec criteria arc satisfied, which adds to the Cisco Unified Border
Element roleof a demarcation point within a VoIP network.
Ifcodec filtering isnot required, aCisco Unified Border Element also supports transparent
codec negotiations. This support enables negotiations between endpoints with the Cisco
Unified Border Element simply by leaving the codec information untouched.
Whether performing codec filtering or operating in transparent mode, aCisco Unified Border
Element must support thecodec that is used bytheendpoints.
Gatekeeper andCisco Unified Border Element Implementation
Cisco Unified Border Element Codec Filtering Examples
This subtopic compares the Cisco Unified Border Element codec filtering methods.
Cisco Unified Border Element Codec
Filtering Examples
*Cisco Unified Border Element codec negotiation:
VoIP 1
VoIP 2
1 G 711a-la*
2 G 729A
3 G 7396
2 G 729A
T"g 7298""
Cisco Unified Border Element with codec transparency:
VolP1
2 G 723A
3 G 729B
Cisco Unitied Border Element
VoIP 2.
2 G729A
3 G 7296
The figure shows how codec negotiation is performed on aCisco Unified Border Element. Two
VoIP clouds need tobe interconnected. In this scenario, both VoIP I and VoIP 2networks have
CJ.711 a-law as the preferred codec.
In the first example, the Cisco Unified Border Element isconfigured touse the (I.729A codec.
Ihis configuration can be done b> simply using the appropriate codec command onboth VoIP
dial peers. When acall is set up. the Cisco Unified Border Element will only accept G.729A
calls, thus influencing thecodec negotiation.
In the second example, the Cisco Unified Border Element is configured lor atransparent codec
and will leave the codec infonnation contained within the call signaling untouched. Because
both VoIP I and VoIP 2have G.711 a-law as their first choice, the resulting call will be a
G.71! a-law call.
5-60 Implementing Cisco Voice Communications and QoS(CVOICEI v80
Configuring Media Flow and Transparent Codec
This topic describes how to configure the available Cisco Unified Border Element media flow
options and codec transparency.
Configuring Media Flow and
Transparent Codec
router(coafig-dial-peer)#
router{conf-voi-servl #
router(config-claBs>#
media [flow-around | flow-through]
Configures media flow-around orflow-through ona dial peer
Available indial-peer, voice service voip, orvoice class
* Media flow-around supported only for SIP-to-SIP or H.323-to-
H.323
Default: flow-through
router (conflg-claas) # ^^^^^^^^^^^^^^^^
codec transparent
Configures transparent codecpass-through in dial-peer or
codec class
TheCisco Unified Border Element media flow and codec transparency can beconfigured using
various configuration elements.
media Command
To configure media flow-through ormedia flow-around, use the media command. This can be
done in dial-peer configuration mode, globally under tlie voice service configuration mode, or
ina voice class that can then bereferenced bymultiple dial peers. Thedefault is media flow-
through. Media flow-through isthe only supported method for 11.323-to-SIP interworking.
codec transparent Command
To configure transparent codec pass-through, use the codec transparent command. This can
be done indial-peer configuration mode or via a codec class.
Media Flow-Around and Transparent Codec Example
This topic presents asample Cisco Unified Border Element configuration for media flow-
around and codec transparent.
Media Flow-Around and Transparent
Codec Example
Cisco Lniêd
Communications
Manager E>p'ess H__5 an_ HmM, M225 and Manager Express
10,11 '1245 SHecofle S1 H225, H245 filte pn_ " ând
RTP Cisco UnrfiedN. RTP Cisco Unified
Boroer Element \ Border Element
192 16B1 1 \ 192 168 2.1
voice service v oip
allow-conoecti Ons h323 to h323
h323
call etart in terwo rk
dial -peer voice 10 V Oip
des;i nation-pa[tern l..
media flOK-aro und
codec transpar ent
sessi tatget ipv4 : 10. 1.1.1
dial-peer voice 20 v oip
deati lation-pa ttern ?..
media lo-ato und
codec ttanspar ent
seem target ipv4 192 .168.2.1
Cisco Unified
Communications
fhe figure illustrates asample Cisco Unified Border Hlcment configuration for media flow-
around and codec transparency. The configuration consists of H.323-to-H.323 signaling
pennission and the respective VoIP dial peers. The dial peers are configured for media flow-
around and codec transparency. These settings can beconfigured inthe voice class and codec-
class and referenced h> the dial peers.
5-62 Implementing CiscoVoice Communications and OoS (CVOICE) vS.O
2010Crsco Systems. Inc
RSVP-Based CAC on Cisco Unified Border
Element
This topic describes how to provide RSVP-based CAC using the Cisco Unified Border
Element.
RSVP-Based CAC on Cisco Unified
Border Element
CiscoUnified Border Element can use standard Cisco IOS
gateway RSVPcall support.
Enables RSVP-based CAC:
- Supportfor voice and videocalls
Requirements:
- Two CiscoUnified Border Elements can be used as RSVP
peers
- Media flow-through toensurethat the reserved path is
used
Because theCisco Unified Border Element is a CiscoIOS gateway, il alsosupports RSVP-
based CAC. Two Cisco Unified Communications Manager clusters caninterconnect using
CiscoUnified Border Element, thusenabling intercluster RSVP-based CAC. RSVP supports
both voice and video calls.
RSVP requires at least two RSVP peers, sotwo Cisco Unified Border Element gateways are
required to enable RSVP-based CAC. When deploying Cisco Unified Border Element and
RSVP-based CAC. you must be sure that the flows that should utilize RSVP are configured for
media flow-through. Media flow-around isnot supported together with RSVP-based CAC.
RSVP-Based CAC
fhis subtopic describes the role ofCisco Unified Border Element in RSVP-based CAC
Cisco Up tied
Communicalions
Manager E-press ^ and
.^C H 245
M225/H 245
. RSVP
Cisco Unite a
Communications
H225 and Manager Express
H 245
Cisco Unifled RTP Cisco Unified
Bordei Border
Element Elemenl
The figure illustrates the placement of tuo Cisco Unified Border Elements to provide RSVP
based CAC. Thecalls are admitted to cross the WAN only when a reservation canbe
succcssfulh made for a call.
5-64 Implementing Cisco VoiceCommunications and OoS (CVOICE] v8 0
RSVP-Based CAC Call Flow
This subtopic describes the call flows in aCisco Unified Border Element deployment with
RSVP-based CAC.
RSVP-Based CAC Call Flow
H.323 Fast Start
14 Ringback
Cisco Unified
Cisco Unified
Border
Border
Elemert
Element
?r..iK*..,iH2V SRRVPPail
RSVP Reservation
| mi " ""
: 5. Call Setup (H.2.45)
p Call Proceeding i , 8.Call Proceeding
13 Alerting (H.245} _ 12. Alarting (H.245)
5 RTP/RTCP StreanU ((tow-through)
6. Call Setup (H24.5)
7 Call Proceeding
11. Alering (H.245)
10. Ring
15. Answer
"The figure depicts the signaling flow with two Cisco Unified Border Elements that provide
RSVP-based CAC and use H.323 fast start on all call legs. The relevant step in this scenario
takes place after the call setup message is received by aCisco Unified Border Element. Before
it fomards the call setup message to the other Cisco Unified Border Element, it cheeks the
required bandwidth. The reservation process involves two messages: RSVP Path message that
is processed bv each router in the path from the originating Cisco Unified Border Element to
the terminating Cisco Unitied Border Element, and the reservation message that flows in the
reverse direction. The Path message carries the request with associated parameters, and the
reservation message is used to commit the reservation on all hops. The originating Cisco
Unified Border Element sends the Call Setup after asuccessful reservation message is received.
Eor RSVP-based CAC. media flow-through must be used toensure that the media packets
actually follow the reserved path. In this example, early media is negotiated that allows the
gateways to establish the media flow before the call is answered.
Gatekeeper andCisco Unified Border Element Implementation
5-65
Cisco Unified Border Element Call Flows
This topic describes typical call Hows in the Cisco Unified Border E
cment.
SIP carrier interworking
H.323-to-SIP
RSVP-based CAC
Two Cisco Unified Border Elements with R323-to-H.323
SIP carrier interworking with gatekeeper-based CAC
H.323-to-SIP
H.323 gatekeeper RAS
Call signaling depends on network topology and features that are implemented on the Cisco
Unified Border Element.
This topic describes call flows lor these Cisco Unified Border Element scenarios:
Cisco Unified Communications Manager Express >Cisco Unified Border Element >SIP
carrier
Cisco Unified Communications Manager Express >Cisco Unified Border Element with
RSVP >Cisco Unified Communications Manager Express
Cisco Unified Communications Manager Express >gatekeeper >Cisco Unified Border
Element > SIP carrier
5-66 Implementing Cisco Voice Communrcalions andQoS (CVOICE) v8 0
SIP Carrier Interworking
This subtopic describes the SIP carrier interworking scenario.
SIP Carrier Interworking
Cisco Unfed
Commumcaions
Manager Express
SIP
Camer
The figure shows asimple Cisco Unified Border Element deployment where the Cisco Unified
Border Element isused totranslate the H.323 call leg with the Cisco Unified CommumcaUons
Manager Express to aSIP call leg point to aSIP carrier. Because this is aconnection to an
external VoIP network, media flow-through isrequired tohide internal IP addresses.
SIP Carrier Interworking Call Flow
This subtopic illustrates the call flows in the SIP carrier interworking scent
scenario.
H.323 Slow Start-to-SIP Delayed Offer
Cisco Unified
Border
Element
Enterprise t.~=^.
' IP
SIP
H225/Q931
Call Setup j
H245 1
Capabilities
Negotiation J
1 Initiate Call
rGS = TermnaiCac
C_C = Open lc^ ca
bUP = Session Des
1 7 Call Proceeding -
10 Alerting
14 Connect
15 TCS
16 Master/Slave
17 OLC
. 3-In*,. ....... j....AJMiteJ. .,,,
6_100Tryinq,_^ [ A.5.J0Q Trying^
13 200 OK,(SDPJ s12JO0.OKt(SDPJ
SIP
18 ACK(SDP) I 19 ACK(SDP)
20 RTP.'RTCP Streams
fruuDii~c.iaiiir ::m
(Onlyflow-through supported)
The figure illustrates the call signaling flow when Cisco Unified Border Element provides
interworking scr\ ice between H.323slowstart andSIPdelaved offer.
SIP Carrier Interworking C<
H.323 Fast Start-to-SIP Early Offer
H 225/0 931
Call Setup
With H 245
Capabilities
Negotiation
1 Initiate Cal
Cisco Unified
Border
Elemenl
...stup (rUfeW -,.3,lQvitg.lspp) i...4-ln.vite.(fiP.K ...
t 7 Proceeding j S6J00 Trying ._. \^5JOT Tryinq. .._.
1 10 Alerting (H 245); J)jeDJin&ingJSDP) j^jMSO Ringing
12 RTP/RTCP Streams (flow-through) I s|p
.*.. J...X.j jwN*. ,""<iii..C.U:iiiiiEiu..m.'a.jjiiiiijijiiJuu.-""i'lni!
15 Connect .14 200OK .13 200 OK
16 ACK '17 ACK
fhe figure illustrates the call signaling flow when Cisco Unified Border Element provides
interworking service between H.323 fast start andSIPearlvoffer.
SIP Carrier Interworking with Gatekeeper-Based CAC Call
Setup
This subtopic describes the call flows in aCisco Unitied Border Element deployment with
gatekeeper-based CAC.
SIP Carrier Interworking with
Gatekeeper-Based CAC Call Setup
H.323 Slow Start-to-SIP Delayed Offer
Zone AGK ITSP GK
H225/Q 931
causemp
17 Ringback
> ARQ
3 LRQ
6 CaDBetuft
13. CallProteefling
12. 100 Trying
H225
RAS
{H.225
RAS
10. Invite
11.100 Trying
14. Ringing
16 Alerting
20 Connect
15 Ringing
19 200OK(SDPJ 18. 200 OK (SDP)
23. ACK (SDP]
21. H.245 Capacity Exchange
24. RTP/RTCP Streams (flow-tri rough)
ARQ =Admission Request ACF =Adrr.ss.ori Confirm, LCF - Location Request, LCF =Location Confirm
The figure shows the signaling flow with two gatekeepers and one Cisco Unified Border
Element, providing gatekeeper-based CAC in combination wilh SIP carrier interworking. The
call flow from the Cisco Unified Communications Manager Express toCisco Unified Border
Element follows the regular H.225 RAS procedure, inwhich ARQs are sent by both gateways
totheir respective gatekeepers. Location Request (LRQ) and Location Confirm (LCF) are
exchanged between the gatekeepers. The Cisco Unified Border Element then connects the
inbound H.323 call leg tothe outbound SIP call leg. This example illustrates H.323 slow start-
to-SIP delayed-offer interworking on Cisco Unified Border Element. Interworking between
different protocols (H.323 and SIP) supports media flow-through only.
Gatekeeperand CiscoUnified BorderElement Implementation 5-69
Configuring H.323-to-H.323 Interworking
This topic describes hou to implement 11.323-to-l 1.323 intenvorking on aCisco Unified Border
Element.
H.323-to-H.323 Configuration Overvi
J Enable H.323-to-H.323 interworking.
7. Enable fast-start-to-slow-start interworking (optional).
; Configure H.323dial peers.
To configure H.323-to-H.323 intenvorking between aCisco Unified Coninuinications Manager
cluster and aCisco \ nilied Communications Manager Express gateway, follow these steps:
Step 1 Enable H,323-to-H.323 intenvorking.
Step2 Configure fast-start-to-s low-start interworking. if desired.
Step 3 Conligure the H.323dial peers on the Cisco Unified Border Element to allowcall
routing between the call legson bothsides ofthe Cisco Unified Border Element.
5-70 Implementing Cisco VoiceCommunicalions and QoS (CVOICE) vS.O
Configuring H.323-to-H.323 Interworking
This topic describes the mandatory settings for H.323-to-H.323 intenvorking on Cisco Unified
Border Element.
Configuring H.323-to-H.323
Interworking
router Icon fig) tt
voice service voip
Enters voice service VoIP configuration mode
Tallow-connections h323 to h323
Enables H.323-to-H.323 interworking
Default: Only POTS-fo-any and any-to-POTS connections are
permitted
H.323-to-H.323 interworking isdisabled by default. Itisenabled using the aHow-connect ions
h323 to h323 command in global voice service configuration mode. By default, only POTS-to-
any and any-to-POTS connections are permitted.
Gatekeeper andCisco Unified Border Element Implementation 5-71
Configuring H.323-to-H.323 Fast-Start-to-Slow-Start
Interworking
This topic describes how to implement ll.323-to-H.323 last-start-to-slow-start interworking
Ciseo Unified Border Element.
Configuring H.323-to-H.323 Fast-
to-Slow~Start Interworkinq
h323
Enters H.323 mode
routerIconf-serv-h323)#
call start {fast slow j interwork}
Forces theH323 gateway touseeither fast-start (H.323 v2) orslow-
start (H 323 v1) procedures for the dial peers using H.323
interwork option allows Cisco Unified Border Element
interoperability between fast-start and slow-start procedures
Caution. Cisco Unified Border Element with this setting will not
originate anyH.323 calls(fast startandslow-start disabled)
Default fast start (H323 v2)
11.323 fast-start-to-slow-start interworking isenabled using the eall start command inh323
configuration mode. The call start command has three options:
Fast: Ihis selection forces the H.323 gateway touse fast start (H.323v2) procedures for tlie
dial peers using 11.323. This is thedefault setting.
Slow: This option makes the H.323 gateway use slow start (H.323vl) procedures for the
dial peers using H.323.
Interwork; This ke\word allows Cisco Unified Border f.lenient interoperability between
fast start and slow start procedures. This option effectively disables the any-to-fl.323
gatewa\ operations onthe Cisco Unified Border Element because the gateway will not
originate am H.323 calls (fast start and slow start arenotenabled).
5-72 Implementing Cisco Voice Communicationsand QoS (CVOICE) vS.O
"*%*
H.323-to-H.323 Interworking Example
This topic provides aCisco Unified Border Element H.323-to-H.323 interworking example
H.323-to-H.323 Interworking Exampi
Cisco Unified
Communications
Manager E-press H ~-c _.. . __
10111 HZ25 """'" 1 ' RL<flrndeB2 H.225
Cisco Unified
Communications
Manager Express
Cisco Unified'
Border Element'
192.1681 1
IP WW
RTP Cisco Unified
Border Element
192.168 2.1
voice service voip
allow-connectione h323 Co h323
t.323
call start interork
I
description To Cisco unified CUE
1
description To Cisco UBH
session target lpv4:13Z.168 .2.1
The figure illustrates asample configuration for Cisco Unified Border Element H.323-to-H.323
interworking. The configuration consists ofthe H.323-to-H.323 signaling permission, fast-start-
to-slow-start activation, and VoIP dial peers responsible for both call legs ofthe Cisco Unified
Border Element.
Configuring H.323-to-SIP Interworking
This topic describes how to implement H.323-to-SIP interworking on Cisco Unified Bordt
Element,
Configuring H.323-to-SiP Interworl
router(conf-voi-serv}#
allow-connections h323 to sip
allow-connections sip to h32 3
* Enables H.323-to-SIP interworking
In one direction only
Two mirrored statements required forbidirectional
interworking
Default: only POTS-to-any and any-to-POTS connections are
allowed
For SIP-to-SIP interworking:
allow-connections sip to sip
H.323-to-SIP interworking is disabled b\ default. It is enabled using the allow-conncctions
h323 to sip command in global voice service configuration mode. By default, only POTS-to-
any and anwo-POTS connections are permitted.
The configuration fbr H.323 and SIP interworking is unidirectional, thus if bidirectional
interworking is required, you need to configure the mirror-matching statement as well. Eor
example, ifbidirectional Ii.323-to-SIP interworking is required, you need to configure allow-
connections h323 tosip as well as allow-connections sip toh323.
SIP-to-SIP intenvorking is enabled similarly, using the allow-conncctions sip to sip command.
5-74 Implementing Cisco Voice Communications andQoS(CVOICE) v80
Configuring H.323-to-SIP DTMF Relay Interworking
This topic describes how to implement H.323-to-SIP DTME interworking on Cisco Unified
Border Element.
Configuring H.323-to-SIP DTMF Relay
Interworking
router(contig-dial-peer)#
dtmf-relay [cisco-rtp] [h245-alphar.UTi.eric] [h245-signal]
[rtp-nte [digit-drop]] [aip-notlfy]
Basic DTMF relay hterworking
- H245alpha-signerf and SIPRTP-NTE (RFC 2833)
- H.245alpha/signal and SIP Notify
dkjit-drop drops incoming in-band DTMF digits when H.323 call leg uses
out-of-band relay (H.245 alpha/signal andSIP RTPWE)
- Prevents sending DTMF n two channels
Configuredon the SIP call leg dial peer
In-band ci*co-rtp, rtp-nta (RFC2833)
Out-of-band h2-a1phanumeric. h246-signal
rtp-nte (RFC 2833)
sip-notify
DTMF intenvorking is asubset ofH.323-to-SIP interworking and supports these DTMF relay
combinations:
H.245 alpha/signal and SIP RTP-NTE (RFC 2833), as afunction ofbasic DTMF
interworking. This method converts an out-of-band DTMF relay method to an in-band
rela>. Its potential issue is that the DTMF digits are transported both in-band and out-of-
band on the H.323 call leg.
Note NTE - named telephony event
11.245 alpha/signal and SIP Notify, as a function ofbasic DTME interworking. This method
converts anout-of-band DTMF relay method to another out-of-band DTMF relay.
G.711 inband DTMF to RTP-NTE. as a function of supplementary DTMF interworking.
Thismethod converts an in-band DTMF relay method to another in-band DTMF relay.
The digit-drop keyword in the dtmf-relay rtp-nte digit-drop command prevents sending both
in-band and out-ofband tones totheH.323 leg. It isconfigured inthe dial peer thatprovides
the SIP call leg for the first DTMF relay method (H.245 alpha/signal and SIP RTP-NTE). It is
useful only ifeither dtmf-relay h245-alphanumeric ordtmf-relay h245-signal is configured
on the H.323 call leg.
The table in the figure provides areview ofin-band and out-of-band DTMF relay methods that
arc supported in H.323and SIP.
Verifying Cisco Unified Border Element
This topic describes hou to verify Cisco Unified Border Element operations.
show commands
show call history voice
show diat-peervoice
* show voip dp connections
debug commands:
' debug voip ipipgw
1 debug cch323 all
debug ccsip messages
debug h22Sasn1
* debug h225 events
debug h245asn 1
* debug h245 events
' debug voip ccapi inout
Ihe figure summarizes the commands that can be used tovcri IV and debug Cisco Unified
Border Element operations. All commands, except the debug voip ipipgw command, are
Upical commands that areknown from traditional H.323 or SIP environments. To sueccssfulK
troubleshoot Cisco Unified Border Element functionality in ll.323-to-SIP interworking
scenarios, both groups of commands are needed (SIP andH.323).
5-76 ImplementingCisco VoiceCommunications and QoS (CVOICE) v8 0
Debugging Cisco Unified Border Element Operations
This subtopic shows how to debug Cisco Unified Border Element operations.
Debugging Cisco Unified Border
Element Operations
jtert debug voip ipip3
/H323/cch323 Bet_pr.i_co<lae_liBti: Firat preferred codec (bytes) -1(201
/E323/cch323_get_per_info: Flo* Node Bet to FLOW .THROUGH
_/B323/ccn323_build_local_enCOaed_fa.tStartOLC9: sroAddrBBS - OXA01O665,
h245 lpoit 0, flow node - 1,
.../H323/cch323_g8naric_c.pBn_logical_chann8l: current codec - 16:20:20
./a323/cch323_reCeive aatStart_cap raapouBe: Send cap ind to peer leg
,/H3 23/cch323_build olc.forccapi: audioFs.stStartArray-0x4J0*57 54
_./B323/cch323_build_olc_ior_ccapi: Char
Logical Channel Number {fd> : 1
logical Channel Number (rev): :
Channel addraBS !fwd/rev):
RTP Channel lfwd/vl :
RTCP Channel (*d/revl :
OoS Capability Iftrd/rav):
Synnetcic Audio Codec:
Symmetric Audio Codec Bytes:
Flow Hode:
Silence Suppression:
el inrc atio
10.1.Z50.102
16764
1676 5
The figure shows sample output from the debug voip ipipgw command. It includes the
description ofthe media flow (flow-through in this example), and lists negotiated parameters,
such as RTP port numbers.
Gatekeeper and CiscoUnified BorderElementimplementation
5-77
Viewing Cisco Unified Border Element Calls
This subtopic explains how to examine VoIP calls transported over the Cisco Unified Border
t. lenient.
Viewing Cisco Unified Border Elei
outer* show call
Telephony call-legs: 0
SIP call-lege: l
B323 call-legs: 1
Call agent controlled call-legs: 0
SCCP call-legs: C
Multicast call-legs: 0
Total call-legs: 2
137C : 163 346116e00ms. 1 .1580 Pid:4O0C2 has
dur 30:00:22 tx : 11 24/2 24 BO rx:112/2050
IP 10.1.2.26:25850 SRTP: off rtt:0ffiB pl:0/0mB lost:0/0/0 del ay:0/0/0n
g729r8 Textfielay: off
media inactive detected:n media contrl rcvd:n/a timestamp:n/a
long duration gall deEected:n long duration call duration:n/a timesta
1010
The show call active voice briefcommand can beused tovalidate that an active cal! has been
established using the H.323-to-SlP interworking procedure. If so, there should beoneSIP and
one H.323 call leg. Additional!), the output displays other information, such aseall duration
and RTP parameters.
5-78 Implementing CiscoVoice Communications and QoS (CVOICE) vS.O
mm
Summary
Summary
Cisco Unified Border Semen! features include protocol
interworking, address hiding, security, video integration, and
CAC.
Cisco Unified Border Element supports conversion offast-
start-to-slow-start signaling methods within thesameprotocol
(H.323 or SIP).
The default mediaflow-through can be changed to flow-
around when interworking within the same protocol.
CiscoUnified Border Element can be deployed in
combination with RSVP.
CiscoUnified Border Element call flows differ depending on
the CAC method in use.
Summary (Cont.)
H.323-to-H.323 interworking allows the configuration of fast-
start-to-slow-start conversion.
H.323-to-SIP interworking can be combined with gatekeeper
CAC on the H.323 side.
Cisco Unified Border Element can pass codec negotiation
transparently and allow the media toflow around without
being handled.
Debug and verification commands display both VoIP call legs
of Cisco Unified Border Element.
Gatekeeperand Cisco Unified Border Element Implementation 5-79
Implementing Cisco Voice Communications and QoS (CVOICE] vB 0 2010 Cisco Systems, Inc
Module Summary
Module Summary
H.323 gatekeepers make H.323 environments more scalable
by providing address resolution, centralized call routing, and
CAC. In addition, H.323 gatekeepersoffer optional features,
suchas call authorization, call management, and bandwidth
management.
Cisco Unified Border Element connects twoVoIP networks
and can provide protocol interworking, address hiding, and
CAC.
This module describes the role of H.323 gatekeepers and Cisco Unified Border Elements in
Cisco Unified Communicalions. Gatekeepers add scalability to11.323 environments by
assuming acentralized call routing, address resolution, and Call Admission Control (CAC)
function. Gatekeeper-based CAC prevents oversubscription ofWAN bandwidth by limiting the
number of H.323 calls into the network. Gatekeeper-based CAC does not provide any dynamic
checks ofthe utilized bandwidth, as is thecase with Resource Reservation Protocol (RSVP).
Cisco Unitied Border Elements fulfill a range of tasks, including signaling interworking
(H.323-to-Session Initiation Protocol [SIP]), media interworking (How-through, flow-around,
and dual tone multifrequency [DTMF]), topology hiding (media flow-through), billing and Call
Detail Record (CDR) normalization, as well as quality ofservice (QoS) and bandwidth
management functions.
5-82 Implementing Cisco Voice Communications and QoS (CVOICE| w8 0 2010 Cisco Systems, Inc
Module Self-Check
solutions arefound intheModule Self-Check Answer Key.
QI) RAS is asubset of the signaling protocol. (Source: Understanding Gatekeepers)
A) H.323
B) SIP
C) H.225
D) H.245
Q2) Zone prefixes are usually used to associate to aconfigured zone. (Source:
UnderstandingGatekeepers)
A) IP addresses
B) gatekeepers
C) area codes
D) endpoints
Q3) Gatekeepers use technology prefixes to route calls when there is no registered
(by agateway) that matches the called number. (Source: Understanding Gatekeepers)
Q4) Agatekeeper has alogical process for call routing that depends on technology and
prefix matching. (Source: Gatekeepers)
Q5) Directory gatekeepers forward __to gatekeepers. (Source: Understanding
Gatekeepers)
Q6) Asingle gatekeeper can manage multiple local and remote . (Source:
Understanding Gatekeepers)
Q7) Gatekeeper configuration steps are done in the configuration mode on Cisco IOS
routers. (Source: UnderstandingGatekeepers)
Q8) A prefix isan optional H.323 standards-based feature that issupported by Cisco
gateways and gatekeepers that enables more flexibility in call routing within an H.323
VoIP network. (Source: UnderstandingGatekeepers)
Q9) Cisco IOS routers can be registered as . with gatekeepers. (Source: Understanding
Gatekeepers)
q 10) The dial peer determines how todirect calls that originate from alocal voice
port into the VoIP cloud tothe RAS session target. (Source: Understanding
Gatekeepers)
Q11) Use the command todisplay registered endpoints ofthe gatekeeper. (Source:
Understanding Gatekeepers)
Q12) Voice quality is regulated by , which limits the number ofcalls that can be active
on a particular link at the same time. (Source: Understanding Gatekeepers)
Q13) Agatekeeper calculates the bandwidth requirement percall using the formula _.
(Source: Understanding Gatekeepers)
>2010 Cisco Systems. Inc Gatekeeper and Cisco Unified Border Element Implementation 5-83
014) The _command allows the gatekeeper to manage the bandwidth limitations within
azone, across zones, and at aper-session level. (Source: Understanding Gatekeepers)
QI5) l'se the command to \eriiy /one bandwidth. (Source: Understanding
Gatekeepers)
016) Cisco Unitied Border Element interconnects multiple VoIP networks by routing calls
between two dial peers. (Source: Examining Cisco Unilied Border Element)
A) POIS
B) VoA'lM
C) VoKR
D) VoIP
Q17) Protocol interworking interconnects VoIP networks, using the same or different _
protocols. (Source: Examining Cisco Unified Border Element)
A) signaling
B) compression
C) codec
D) transport
018) Which ofthese has the two correct options for directing media streams in aCisco
Unilied Border Element? (Source: Examining Cisco Unified Border Element)
A) b\pass. flow-around
R) flow-through, bypass
C) flow-through, traverse
D) flow-through, flow-around
Q19) Ifcodec filtering is not required, aCisco Unified Border Element also supports _
codec negotiations. (Source: Examining Cisco Unified Border Element)
A) multiple
B) null
C) dynamic
D) transparent
Q20) Which method should beused when deploying Cisco Unified Border Element and
RSVP-based CAC? (Source: Examining Cisco Unified Border Element)
A) media tlow-around
B) media bypass
C) media flow-through
D) media pass-through
E) media pass-\ia
Q21) What is required lointerconnect two Cisco Unified Communications Manager clusters
using RSVP-based CAC' (Source: Examining Cisco Unified Border Element)
A) tuo Cisco Unified Border Elements, media flow-through
B) one Cisco Unified Border Element, media flow-through
C) two Cisco Unified Border Elements, media flow-around
D) one Cisco Unified Border Element, media flow-around
5-84 Implementing Cisco Voice Communications and QoS fCVOICE] v8.0 2010 Cisco Systems Inc
Q22) When Cisco Unified Border Element is used to translate an H.323 call leg with the
Cisco Unified Communications Manager cluster to aSIP call leg point to aSIP carrier.
the call flow must Cisco Unified Border Element. (Source: Examining Cisco
Unified Border Element)
A) bypass
B) flowaround
C) stop at
D) flow through
Q23) The configuration for H.323-to-SIP interworking is _. (Source: Examining Cisco
A) unilateral
B) bilateral
C) unidirectional
D) bidirectional
Q24) Which command is used to enable H.323-to-H.323 interworking? (Source: Examining
Cisco Unified Border Element)
A) allow-connectionsh323 to sip
B) allow-connections h323 to h323 interworking
C) allow-connections h323 interworking
D) allow-connections h323 to h323
Q25) ACisco Unified Communications Manager cluster needs lo route outbound calls using
H.323 toaSIP carrier. Which configuration is required? (Source: Examining Cisco
A) allow-connections h323to sip
B) allow-connections sip to h323
C) allow-connections sip to sip
D) allow-connections h323 tosip and allow-connections sip toh323
026) Which command is used to configure codec pass-through? (Source: Examining Cisco
A) codec pass-through
B) codec transparent
C) codec auto
D) codec flow-through
Q27) Which command is used toverify Cisco Unified Border Element operations? (Source:
Examining CiscoUnifiedBorder Element)
A) debug h323 events
B) debug ras
C) debug voip cube
D) debug voip ipipgw
i2010 Cisco Systems. Inc. Gatekeeper and Cisco Unified Border Element Implementation 5-85
on C
Q2t C
<ih F 164 address
Q->) zone
0?) Location Requests(I RQsl
06) /one:,
Q7) tratekeeper
ON) "technolog "or "teeli-"
09) yatt;wa\ v
QIO) VoIP
QUI show gatekeeper endpoints
Q12)
Call Admission Control (CAC)
QI3) Per-call bandwidth -= codec bandwidth
014) band" idth
015) show gatekeeper zone status
016 j I)
Ql-i A
QiSi 13
Qiyi 1)
Q20) C
02D
A
02:i D
Q2.') C
Q24i D
Q25) A
Q26i li
Q2?) D
Module 6
Quality of Service
Overview
Converged networks must be engineered properly to guarantee satisfactory VoIP service. This
module describes quality of service (QoS) requirements; conceptual models such as best-effort.
Integrated Services (IntServ), and differentiated services (DitTServ); and the implemental.on ol
QoSon Cisco IOSplatforms.
The module covers the theorv ofQoS. design issues, and configuration ofvarious QoS
mechanisms to facilitate the creation ofeffective QoS administrative policies, with aspecial
focus on voice transport. It provides design and usage rules for various advanced QoS features
and for the integration of QoS with underlying Layer 2QoS mechanisms. The module enables
learners to design and implement efficient, optimal, and trouble-free multiservice networks that
guarantee satisfactory voice quality.
Module Objectives
Upon completing this module, you will be able to describe why QoS is needed, what functions
it performs, and how it can be implemented in aCisco Unified Communications network. Ihis
ability includes being able tomeet these objectives:
Hxplain the functions, goals, and implementation models of QoS, and what specific issues
and requirements exist in aconverged Cisco Unified Communications network
Describe the characteristics and QoS mechanisms oftheDiffServ model and contrast itto
other models
Explain the operation and configuration ofthe QoS classification and marking mechanisms,
including the concept oftrust boundaries and describe how I.FI and cRTP provide link
efficiency onWAN links and how they are configured
Explain policing, shaping, and LLQ, their operations and configuration, using the MQC
Describe how AutoQoS works and what itachieves in aCisco Unified Communications
network
Lesson 1
Introducing QoS
Overview
IP networks must provide anumber of services to adequately support voice transmission using
VoIP. These services include security, predictability, measurability, and some level ofdelivery
guarantee. Network administrators and architects achieve this service level by managing delay.
delay variation (jitter), bandwidth provisioning, and packet loss parameters with quality of ^
service (QoS) techniques. This lesson introduces the concept ofaconverged network, identifies
four problems that could lead to poor quality ofservice, and describes solutions to those
problems. It also explains and evaluates the three generic models of implementing QoS.
Objectives
Upon completing this lesson, you will be able to describe the goals, functions, and
implementation models ofQoS, and explain the issues and requirements that need to be
addressed for voice and video transmission. This ability includes being able tomeet these
objectives:
Explain the four key quality issues for voice traffic that exist in Cisco Unified
Communications networks anddescribe howthey impact voice quality
Define QoSgoals withrespect to voice traffic
Explain the three key steps that are involved in implementing aQoS policy in aCisco
Unitied Communications network
Describe howtraffic is identified anddivided intoclasses, andhowQoS policies are
defined for tlie traffic classes
I.ist four methods for implementing and managing a QoS policyCLI, MQC, Cisco
AutoQoS. andQPMand describe theircharacteristics
Describe briefly the three key models for providing QoS ina network
QoS Issues
This topic descnbes typical traffic patterns ina traditional data network.
rqec
Traditional data traffic characteristics:
Bursty dataflow
First-come, first-served access
* Mostly not time-sensitivedelays acceptable
Brief outages are survivable
Main Campus
Before networks converged, network engineering was focused on connectivity. The rates at
which data came onto the network resulted in bursty data flows. Data, arriving in packets, tried
to grabas muchbandwidth as it couldat any giventime. Access was on a first-come, first-
served basis, Ihe data rate available toany one user varied, depending on the number of users
accessing the network at any given time.
The protocols that have been developed have adapted tothe bursty nature ofdata networks, and
brief outages are survivable. For example, when you retrieve email, adelay ol'afew seconds is
generally not noticeable. Adelay of minutes is annoying but not serious.
Iraditional networks also had requirements for applications such as data, video, and Systems
Network Architecture (SNA). Since each application hadditTerent trafficcharacteristics and
requirements, network designers deployed nonintegrated networks. These nonintegrated
networks weredesigned to carrya specifictype of traffic: data network. SNAnetwork, voice
network, and \ideo network.
ImplementingCisco VoiceCommunicationsand QoS (CVOICE) v8.0
2010Cisco Systems Inc
After Converged Networks
This subtopic explains the characteristics of various traffic types in aconverged network that
transports both data and time-sensitive traffic such as voice and video.
After Converged Networks
Remote Campus
Lmimwmm-.
Constant smait-packet voice flow
competes withburstydata flow
Criticaltrafficmust get priority
Voice and video am time-sensitive
Bnef outages not acceptable
Tetophony
Main Campus
Problemexample
1 cannot understand you; your
voiceis breaking up,"
Teleconferencing "The picture is very jerky. Voice is
not synchronized."
Call Center "Please IWtowhilemyscreen
refreshes."
The figure illustrates aconverged network in which voice, video, and data traffic use the same
network facilities.
Although packets carrying voice traffic are typically small, they cannot tolerate delay and delay
variation as they traverse the network. Voices will break up and words will become
incomprehensible.
On the other hand, packets carrying file transfer data are typically large and can survive delays
and drops. It is possible to retransmit part ofadropped data file, but it is not feasible to
retransmit a part of a voice conversation.
The constant, small-packet voice flow competes with bursty data flows. Unless some
mechanism mediates the overall flow, voice quality will be severely compromised attimes of
network congestion. The critical voice traffic must get priority. Voice and video traffic is very
time-sensitive. Itcannot be delayed and it cannot be dropped, orthe resulting quality of voice
and video will suffer.
Finally, converged networks must not fail. While afile transfer or email packet can wait until
the network recovers, voice and video packets cannot wait. Even a brief network outage ona
converged network can seriously disrupt business operations. With inadequate preparation of
the network, voice transmission ischoppy orunintelligible. Gaps in speech are particularly
troublesome when pieces ofspeech are interspersed with silence. In voice-mail systems, this
silence is aproblem. For example, when 68614 is dialed and the gaps in speech are actually
gaps in the tone. 68614 becomes 6688661144, because the gaps in speech are perceived as
pauses in the touch tones.
Quality of Service
Quality Issues in Converged Networks
This subtopic describes QoS issues in a converged IP network.
Quality issues in Convert
' Lack ofbandwidth: Multiple flows compete fora limited
amount of bandwidth.
End-to-end delay (fixed and variable): Packets have to
traverse many network devicesandlinks that add up tothe
overall delay.
Variation of delay (jitter): Sometimes there is much other
traffic, which results inmore delay.
Packet loss: Packetsmay have tobe dropped when a link is
congested.
The four major problems facing converged enterprise networks include the following:
Bandwidth capacity: large graphics Hies, multimedia uses, and increasing use ofvoice
andvideo cause bandwidth capacity problems overdata networks.
End-to-end delay (both fived and variable): Delay is the time that it takes for apacket to
reach the receiving endpoint alter being transmitted from the sending endpoint. fhis period
of time is called "end-to-end delay." and consists of two components:
Fixed network delay: Two tvpes offixed delays are serialization and propagation
delav s. Serialization is the process ofplacing bits on the circuit. The higher the
circuit speed, the lesstime it takes toplacethe hitson thecircuit. Therefore, the
higher the speed ofthelink, the less serialization delay is incurred. Propagation
delav is thetime that it takes for frames to transit thephysical media.
Variable network delay: Processing delay isa type of variable delay, and isthe
time that is required by a networking device tolook up the route, change the header,
and complete other switching tasks. In some cases, thepacket must also be
manipulated, as. for example, when the encapsulation type orthe hop count must be
changed. Fiach ofthese steps can contribute tothe processing delay.
Variation of delay(also called jitter): Jitter isthedelta, or difference, in thetotal end-lo-
end delav \ aluesof twovoicepacketsin the voice flow.
Packet loss: Loss ofpackets is usually caused by congestion in the WAN, resulting in
speech dropouts orastutter effect ifthe playout side tries loaccommodate by repeating
previous packets.
Lack of Bandwidth
This subtopic explains how to identify the lack ofbandwidlh in aconverged network.
Lack of Bandwidth
Maximum available bandwidth equals the bandwidth of the weakest link.
. Multiple flows are com peting for ttie same bandwidth, resulting in much
less bandwidth being availableto one singleapplication.
IP
_\ hsekb-sp h^^s H I VI
^J 1 ' ' ' 1100 urn p
Bandwidth maamum =minimum of (10 Mb/s, 256 kb/s, 512 kb/s. 100 Mb/s) =256 kb/s
Bandwidth available =bandwidth maximum/flows ^^
The figure illustrates an empty network with four hops between aserver and aclient. Each hop
is using different media with adifferent bandwidth. The maximum available bandwtdth is equal
to the bandwidth ofthe weakest (slowest) link.
The calculation ofthe available bandwidth, however, is much more complex in cases where
multiple flows traverse the network. The calculation ofthe available bandwidth in the
illustration is an approximation.
Quality of Service
Managing Available Bandwidth
This subtopic explains the methods lo address the lack of bandwidth in aconverged network.
Ways to Increase or
Ban
Upgrade the linkthe best solution but also the most expensive.
Forwardthe important packets first
Compress the payload ofLayer 2frames (it lakes time].
Compress IP packet headers
ihe best way to increase bandw idth is to increase the link capacity to accommodate all
applications and users, with some extra bandwidth tospare. Although this solution sounds
simple, increasing bandwidth is expensive and lakes time toimplement. There are often
technological limitations in upgrading toa higher bandwidth.
Another option is to classify traffic into QoS classes and prioritize traffic according to
importance. Voice and business-critical traffic should gel sufficient bandwidlh to support their
application requirements, voice should get prioritized forwarding, and the Icasl-important
traffic should get whatever unallocated bandwidth isremaining. Avariety ofmechanisms such
as these are av ailable in Cisco IOS QoS Software to provide bandwidth guarantees:
Priority queuing (PQ) or custom queuing (C'Q)
Modified deficit round robin (MDKR) (on Cisco 12000 Series Routers)
Distributed tv pe ofservice (ToS)-hased and QoS group-based weighted fair queuing
(WFQ) (on Cisco 7.v00 Series routers)
Class-based weighted fairqueuing (CBWFQ)
Low latency queuing (LLQ)
Optimizing link usage by compressing the payload offrames (virtually) increases the link
bandwidth. Compression, on the other hand, also increases delay because ofthe complexity of
compression algorithms. Using hardware compression can accelerate packet payload
compressions. Stacker and Predictor are two compression algorifhms that arc available in Cisco
IOS Software,
Another link efficiency mechanism is header compression. Ileader compression is especially
effective in networks in which most packets carry small amounts ofdata (that is. where
pav load-to-header ratio is small), Fvpical examples of header compression are TCP header
compression and Real-Time Transport Protocol (RTP) header compression.
6-8 Implementing Cisco Voice Communications andQoS(CVOICE) vS.O
2010 Cisco Systems, (no
End-to-End Delay
This subtopic explains the components ofend-to-end delay.
End-to-End Delay
End-to-end delay equals a sum of all propagation, processing, and queuing
delays in the path.
In best-effort networks, propagation delay isfixed, and processing and
queuing delays are unpredictable.
Delay =P1 +Q1 +P2+Q2 +P3+Q3 +P4=Xms
The figure illustrates the impact that anetwork has on the end-to-end delay of packets going
from one end ofthe network to the other. Hach hop in the network adds to the overall delay as
follows:
Propagation delay is caused by the speed of light traveling in the media (fiber optics or
copper media).
Serialization delay is the time that it takes to clock all the bits in apacket onto the wire.
This is a fixed valuethat is a function ofthe linkbandwidth.
fhere areprocessing andqueuing delays within a router.
Propagation delay is generally ignored but it can be significant. It amounts to about 40 ms
coast-to-coast, over optical fiber.
Example: Effects of Delay
Acustomer has arouter in New York and arouter in San Francisco, each connected by a 128-
kb/s WAN link. The customer sends a 66-byte voice frame. To transmit the frame (528 bits), it
will take 4.125 ms to clock out (serialization delay). However, the last bit will not arrive until
40 ms after itclocks out (propagation delay). The total delay equals 44.125 ms.
This calculation will bedifferent ifthe circuit ischanged toaT1. To transmit the frame (528
bits), itwill take 0.344 ms to clock out (serialization delay). However, the last bit will not arrive
until 40 ms after transmission (propagation delay) for a total delay of40.344 ms.
Quality of Service 6-9
Types of Delay
This subtopic describes tlie various types ofdelay.
Processing Delay The lime to take the packet from the input interface examine it
and put it intothe output queue
Queuing Delay The time a packet isheld in the output queue
Serialization Delay The timeto place the 'bits on the wire"
Propagation Delay: The time ittakestotransmit a packet
Propagation Delay
In general, thereare four typesof delav. as follows:
Processing dela\: The time il takes for arouter to take the packet from an input interface
and put the packet into the output queue ofthe output interface. The processing delay
depends on these factors:
CPl' speed
CPU utilization
IP switching mode
- Router architecture
Configured features onboth input and output interfaces
Queuing delay: The time apacket resides in the output queue ofarouter. Queuing delay
depends on the number ofand sizes ofpackets already in the queue, the bandwidth ofthe
interface, and the queuing mechanism.
Serialization delay: The time it takes to place a frame on ihe physical medium for
transport.
Propagation dela>: The time it takes lo transmit apacket, which usually depends on the
type of media interface.
6-10 Implementing Cisco Voice Communications and QoS(CVOICE| vS.O
Reducing Delay
This subtopic explains the methods to reduce delay.
Reducing Delay
Upgrade theInk (test but most expensive solution).
Forward the important packetsfirst.
Compress thepayload of Layer 2 frames.
Compress IP packet headers
Assuming that arouter is powerful enough to make aforwarding decision rapidly, most
processing, queuing, and serialization delay is influenced by the following factors:
Average length ofthe queue
Average length of packets in thequeue
Link bandwidth
These approaches allow you to accelerate packet dispatching ofdelay-sensitive flows:
Increasing; link capacity: Sufficient bandwidth causes queues to shrink so that packets do
nol wait long before transmittal. More bandwidth reduces serialization lime.
Prioritizing delay-sensitive packets: This isamore cost-effective approach. PQ, CQ,
strict-priority, or alternate-priority queuing within the MDRR (on Cisco 12000 Series
Routers), and LLQ each have preemptive queuing capabilities.
Compressing the payload: Payload compression reduces the size ofpackets, thereby
virtually increasing link bandwidth. Compressed packets are smaller and take less time to
transmit. Compression uses complex algorithms that take time and add todelay. This
approach is not used toprovide low-delay propagation ofpackets.
Compressing the packet header: Header compression is not as CPU intensive as payload
compression and you can use itwith other mechanisms toreduce delay. Header
compression is especially useful for voice packets that have abad payload-to-header ratio,
which you can improve by reducing the header ofthe packet (RTP header compression).
Packet Loss
"fhis subtopic explains how packet loss affects VoIP quality.
Tail drops occur when the output queue isfull These arecommon drops
whichhappen when a link is congested
Many other types of drops exist, usually the result of router congestion,
that are uncommon and may require a hardware upgrade (input drop,
ignore overrun frame errors)
fhe usual packet loss occurs when routers ntn out ofbuffer space for aparticular interface
output queue. The figure illustrates a full interface output queue, which causes newly arriving
packets to be dropped, fhe tenn that is used for such drops is simply "output drop"' or"tail
drop" (packets aredropped at thetail ofthe queue).
Routers might also drop packets for these other less common reasons:
Input queue drop: The main CPU is congested and cannot process packets (the input
queue is full).
Ignore: fhe router ran out of buffer space.
Overrun: fheCPU iscongested and cannot assign a free buffer tothe new packet,
Frame errors: There isa hardware-detected error in a framecyclic redundancy check
(CRC). runt, or giant.
Preventing Packet Loss
This subtopic describes the ways to prevent packet loss.
Preventing Packet Loss
Upgrade the link (best but most expensive solution).
Guarantee enough bandwidth to sensitive packets.
Prevent congestion by randomly dropping less-important packets before
congestion occurs
Custom Queuing (CO)
Modified Daflelt Round Robin(MDRR)
Ctass-Based Weighted Fair Queuing (CBWFQ)
Packet loss is usually the result of congestion on an interface. Most applications that use TCP
experience slowdown because I'CP adjusts to the network resources.
You can follow these procedures to prevent drops ofsensitive applications:
Increase link capacity toease orprevent congestion.
Guarantee enough bandwidth and increase buffer space to accommodate bursts of fragile
applications.
Prev ent congestion by dropping other packets before congestion occurs. You can use
weighted random early detection (WRED) to start dropping other packets before
congestion occurs.
Quality of Service
QoS and Voice Traffic
This topic deseribes the role of QoS in the enterprise strategy to ensure agood
Theability ofthe network to provide
better or "special" service to a set of
users and applications to the detriment
of other users and applications
Voice - Video - Data
#
Consistent,
Predictable Performance
voice service.
QoS is the abilitv ofthe network to provide better or -'special" service to selected users and
applications, to the detriment ofother users and applications.
Cisco IOS QoS features enable vou to control and predlclably service avarietv ofnetworked
applications and traffic tvpes. which enables you to take advantage of anew generation of
media-rich andmission-critical applications.
fhe goal of QoS is to prov ide better and more predictable network service by providing
dedicated bandwidth, controlled jitter and latency, and improved loss characteristics. QoS
achieves these goals by providing tools for managing network congestion, shaping network
traffic, using expensive wide-area links more efficiently, and setting traffic policies across the
network, QoS offers intelligent network services that, when correctly applied, help to provide
consistent, predictable performance.
QoS Policy
This subtopic explains how QoS policy defines forwarding behavior for various traffic types.
QoS Policy
Anetworkwide definition ofthe
specific levels ofquality ofservice
assigned todifferent classes of
network traffic
ABCCorporation example:
ERP
:150ms SRTPovsr M-F
WAN
High
Manutaatiring High
traffic
HTTP/HTTPS Low
Encrypt
Cleaitext
365x24
M-F
HTTPProxy M-F6-10pm
ERP =Enterpnse resource planning
SRTP= Secure Real-Time TransportProtocol
ABC Corporation
Network QoS Policy
Voice Traffic
Absolute Priority
ERP System
Critical Priority
Manufacturing System
Critical Priority
Net Surfing
Not allowed during
business hours
AQoS policy is anetworkwide definition ofthe specific levels of QoS assigned to different
classes of network traffic.
In aconverged network, having aQoS policy is as important as having asecurity policy. A
written and public QoS policy allows users to understand and negotiate lor QoS tn the network.
The figure illustrates asample QoS policy for an organization.
QoS for Unified Communications Networks
This topic describes the implementation of QoS in Cisco Unified Communications networks.
QoS for Cisco Unified Communication:
irl
Identify traffic and ifs
requirements
Divide traffic into classes
Define QoS policies for
each class
Voce Mission-
Critical
DATA
follow these three basic steps toimplement QoS ona network:
Stepl
Identify traffic and its requirements. Study the network to determine the type of
traflic running on the network and then determine the QoS requirements for the
different tvpes of traffic.
Group the traffic into classes with similar QoS requirements. For example, four
classes oftraffic can be defined: voice, high priority, low priority, and browser.
Define QoS policies that will meet the QoS requirements for each traffic class.
Step 2
Step 3
Example: Three Steps to Implementing QoS on a Network
In atypical network, voice will always require absolute minimal delay. Some data that is
associated vsith kev applications will require very low delav (transaction-based data that is used
in airline reservations or online banking applications). Other types ofdata can (olerate agreater
amount of delav (tile transfers and email). Nonbusiness network surfing can also be delaved or
even prohibited.
Aone-to-one mapping between traflic classes and QoS policies is not necessary, for example
three QoS policies could be implemented to meet the requirements ofthe four traffic classes
that are defined inthe example:
NoDelay: Assign to voice traffic
BestService: Assign tohigh-priority traffic
Whenever: Assign toboth the low-priority and browser traffic
Step 1: Identify Traffic and Its Requirements
This subtopic describes the first step in QoS implementation, which is lo identify traffic types
and their requirements.
Step 1: Identify Traffic and its Requirements
Network audit
- Identify traffic on the
network
Business audit
- Determine how each type
of traffic is important for
business
Service levels required
- Determine required
response time
Network
Statistics
The first step in implementing QoS is identifying the traffic on the network and determining
QoS requirements for the traffic. Anetwork audit is recommended because many enterprises
have afalse idea ofwhat applications are running in their networks. IfQoS mechanisms arc
deployed based on an unrealistic baseline, unexpected results may occur.
The next step is determining the QoS problems ofusers. Measure the traffic on the network
during congested periods. Conduct CPU utilization assessment on each oftheir network devices
during busy periods to determine where problems might be occurring.
Next, determine the business model and goals, and obtain a listof business requirements, in
order todefine the number ofclasses sothat you can determine the business requirements for
each traffic class.
Finally, define the service levels that are required by different traffic classes in terms of
response time and availability.
Step 2: Divide Traffic into Classes
This subtopic describes the second step in QoS implementation, which is lo divide traffic into
clas.ses.
Email
Application
Traffic
E-Commerce.
Web Browsing
Voice
Differentiated IP Services
Traffic
Classification
Voice Low Latency
Guaranteed
Transactional Guaranteed Delivery
No Delivery Guarantee
After you have identified and measured the majority ofnetwork traffic, you can use the
business requirements to define traffic classes.
Because ofits stringent QoS requirements, voice traffic will almost always exist in aclass by
itself Cisco has developed specific QoS mechanisms, such as LI,Q, that ensure that voice
aluay s receives priority treatment over all other traffic.
Atier you define the applications with the most critical requirements, you can define the
remaining traffic classes using thebusiness requirements.
Example: Traffic Classification
A typical enterprise might define five traffic classes as follows:
Voice: Absolute priority for VoIP traffic
Mission-critical: Small set of locally defined critical business applications
Iransactional: Database access, transaction services, interactive traflic. preferred data
services
Best-effort: Internet, email
Scavenger (less-than-best-eflbrt): Napster. Kazaa. and olher point-to-point applications
6-18 Implementing Cisco VoiceCommunicationsand QoS (CVOICE) v8
Step 3: Define Policies for Each Traffic Class
This subtopic describes the third step in QoS implementation, which is the definition of policies
for each traffic class.
Step3: Define Policies for Each Traff
Set minimum bandwidth guarantee.
Set maximum bandwidth limits.
Assign priorities to each class.
Manage congestion.
Finally. define aQoS policy for each traffic class, which involves these activities:
Set a minimum bandwidth guarantee
Set a maximum bandwidth limit
Assign priorities to each class
Use QoS technologies, such asadvanced queuing, tomanage congestion
Example: Defining QoS Policies
Using the traffic classes that were previously defined, you can determine QoS policies as
follows:
Voice: Minimum bandwidth: 1Mb/s. UseQoS marking to mark voice packets as priority
level 5: use LI.Q to always give voice priority.
Mission-critical: Minimum bandwidth: 1Mb/s. UseQoS marking tomarkcritical data
packets as priority level 4: use CBWFQ to prioritize critical class traffic flows.
Best-effort: Maximum bandwidth: 500kb/s. UseQoS marking to markihese datapackets
as priority level 2: use CBWFQ toprioritize best-effort traffic flows that are below
mission-critical and voice.
Scavenger: Maximum bandwidth: 100 kb/s. Use QoS marking tomark Icss-than-best-
effort (scavenger) data packets as priority level 0; use WRED todrop these packets when
the network has a propensity for congestion.
QoS Requirements
This topic delines the QoS requirements for voice and video transmission in packet networks.
Latency < 150 ms*
Jitter < 30 ms*
Loss < 1 %*
17-106 kb/s guaranteed
priority bandwidth per call
150 b/s (+ Layer 2 overhead)
guaranteed bandwidth for
voice-control traffic per call
' One-way requirements
Smooth
Benign
Drop-Sensitive
Delay-Sensitive
UDP Priority
Voice traffic has extremely stringent QoS requirements. Voice traffic usually generates a
smooth demand on bandwidth and has minimal impact on other traffic aslong asthe voice
traffic is managed.
While voice packets are typically small (60 to 120 bytes), they cannot tolerate delay ordrops.
The result ofdelay s and drops are poor, and often unacceptable, voice quality. Because drops
cannot be tolerated. User Datagram Protocol (UDP) isused to package voice packets because
TCP retransmit capabilities have no value.
Voice packets can tolerate no more than a 150-ms delay (one-way requirement) and less than I
percent packet loss.
Atypical voice call will require 17 to 106 kb/s ofguaranteed priority bandwidth plus an
additional 150 b/s per call for voice-control traffic. Multiplying these bandwidth requirements
times the maximum number ofcalls that arc expected during the busiest lime period will
prov idean indication ofthe overall bandw idth that is required for voice traffic.
6-20 ImplementingCrscoVoiceCommunicationsand QoS (CVOICE) vB.O
QoS Requirements: Video Telephony
"fhis subtopic defines the QoS requirements for video transmission in packet networks.
QoS Requirements: Video Telephony
Latency < 150 ms*
Jitter < 30 ms*
Loss<1%*
Minimum priority bandwidth
guarantee required is:
- Video stream + 20%
- For example, a 384-kb/s
stream would require 460
kb/s of priority bandwidth
One-way requirements
**"
Bursty
Greedy
Drop-Sensitive
Delay-Sensitive
UDP Priority
Videoconferencing applications alsohave stringent QoS requirements similar to voice. But
videoconferencing traffic isoften bursty and greedy innature and, as a result, can impact other
traffic. Therefore, it is important to understand the videoconferencing requirements for a
network and to provision carefully for it.
The minimum bandwidth for a videoconferencing streamwouldrequirethe actual bandwidth of
the stream (dependent upon thetype of videoconferencing codec being used) plus some
overhead. For example, a 384-kb/s video stream would actually require a total of 460 kb/s of
priority bandwidth.
QoS Requirements: Data
This subtopic contrasts voice and video requirements todata transmission in packet networks.
6-22
Different applications have
different traffic characteristics.
Different versions of the same
application can have different
traffic characteristics
Classify data into relative-priority
model with no more than four to
five classes'
Mission-critical: Locally defined
critical applications
Transactional Interactive
traffic, preferred data service
Best-effort. Internet, email,
unspecified traffic
Less-than-best-effort
(Scavenger): Napster. Kazaa,
peer-to-peer applications
Data
Smooth or Bursty
Benign or Greedy
Drop-Insensitive
Delay-Insensitive
TCP Retransmits
The QoS requirements for data traflic vary greatly. Different applications may make very
different demands onthenetwork (forexample, a human resources application versus an
automated teller machine application). Fven different versions ofthe same application may
have varying network traffic characteristics.
While data traffic can demonstrate either smooth or bursty characteristics, depending upon the
application, data traflic differs from voice and video interms ofdelay and drop sensitivity.
Almost all data applications can tolerate some delay and generally can tolerate high drop rates.
Because datatraffic cantolerate drops, theretransmit capabilities of TCP become importanl
and. as a result, many data applications use TCP.
Inenterprise networks, important (business-critical) applications are usually easy to identity.
Most applications canbe identified based on TCP or UDP port numbers. Some applications use
dynamic port numbers that, to some extent, make classifications more difficult. Cisco IOS
Software supports Network-Based Application Recognition (NBAR), whichyou can use to
recognize dynamic port applications.
It is recommended that data traffic is classified into no more than four to five classes, as
described in the tigure. fhere will still remain additional classes for voice and video.
Implementing Cisco Voice Communications and OoS (CVOICE) v8.i
Methods for Implementing QoS Policy
This topic describes the four methods for implementing an enterprise QoS policy.
Methods for Implementing QoS Policy
Command-line interface (CLI)
Modular QoS CLI (MQC)
AutoQoS
- AutoQoS VoIP (voice QoS)
- AutoQoS for the Enterprise
(voice, video, and data QoS)
QoS Policy Manager (QPM)
- Suite of management
functions
- Enables networkwide QoS
- Monitoring and reporting
Initially, the only way toimplement QoS ina network was by using the command-line interface
(CLI) to individually configure QoS policies at eachinterface. Thiswasa time-consuming,
tiresome, and error-prone taskthat involved cutting and pasting configurations from one
interface to another.
Cisco introduced the Modular QoS CLI (MQC) in order to simplify QoS configuration by
making configurations modular. Using MQC, youcan configure QoS ina building-block
approach, using a single module repeatedly toapply policy to multiple interfaces.
Cisco AutoQoS represents innovative technology thatsimplifies the challenges of network
administration by reducing QoS complexity, deployment time, andcost toenterprise networks.
CiscoAutoQoS incorporates value-added intelligence inCisco IOS Software andCisco
Catalyst software to provision andassist in themanagement of large-scale QoS deployments.
CiscoAutoQoS is an intelligent macrothat enablesyouto enter one or two simpleCisco
AutoQoS commands lo enableall the appropriate features for the recommended QoSsettingfor
anapplication on a specific interface. There are twoversions of CiscoAutoQoS: Cisco
AutoQoS VoIP and AutoQoS for the Enterprise.
QoScan be easilyprovisioned and managed by usingCiscoAutoQoS together with
CiscoWorksQoS Policy Manager(CiscoWorks QPM). CiscoAutoQoS providesQoS
provisioning for individual routers andswitches, simplifying deployment andreducing human
error. CiscoWorks QPM provides centralized QoS design, administration, and traffic
monitoring that scales to large QoS deployments.
Implementing QoS Traditionally Using CLI
This subtopic explains thetraditional method of implementing QoS, the nonmodular CLI,
Implementing QoS Trad ith
Traditional method
Nonmodular
Cannot separate
traffic classification
from policy definitions
Used to augment or
fine-tune newer
AutoQoS method
interface Multllinkl
ip address 10.1.61.1 255.255.255.0
ip tcp header-compression iphc-format
load-interval 3 0
custom-queue-list 1
ppp multilink
ppp multilink fragment-delay 10
ppp multilink interleave
multilink-group 1
ip rtp header-compression iphc-format
1
CLI was the first method to implement QoS ina network. It wasa painstaking task, involving
copying one interface configuration, andthen pasting tl into other interface configurations, CL
took much time and patience.
Ihe original CLI method was nonmodularthere was no wayto separatethe classification of
traffic from the actual definition of policy. You hadtodo both on every interface. Thefigure
illustrates anexample ofthe complex configuration tasks that areinvolved inusing CLI.
WhileCLI is not recommended for implementingQoS policy, il is still used to fine-tune QoS
implementations that have been generated using the Cisco AutoQoS macro.
Implementing Cisco Voice Communications and OoS (CVOICE) vS.O 2010 Cisco Systems. Inc
_m
Implementing QoS with MQC
This subtopic explains theMQC, which provides a scalable approach toimplement QoS.
Implementing QoS with MQC
Acommand syntax for
configuring QoS policy
Reduces configuration
steps and time
claaa-map VolP-ETP
natch accaaa-group 100
clasa-map VoIP-Control
match accaaa-group 101
policy-map QoS-Policy
Configure policy, not "raw"
per-interface commands
Uniform CLI across major
Cisco IOS platforms
clase VolP-RTP
priority 100
claas voiP-Control
bandwidth B
class class-default
Uniform CLI structure for
ail OoS features
Separates classification
engine from the policy
fair-queue
interface aerial 0/0
BBrvica-policy output QoS-Policy
accesa-list 100 permit ip any any
precedence S
accesa-list 100 permit ip any any deep ef
access-list 101 permit tcp any heat
10.1.10.20 range 2000 2002
access-list 101 permit tcp any host
10.1.10.20 range 11000 11999
The MQCis a CLI structure that allows you to create traffic policies and then attach these
policies tointerfaces. Atraffic policy contains oneor more traffic classes and oneor more QoS
features. A traffic class is used lo classify traffic. The QoS features in the traffic policy
determine how to treat the classified traffic.
The MQC offers significant advantages over the legacyCLI method for implementing QoS. By
usingMQC. you can significantly reducethe time and effort that it takes to configure QoSon a
complex network. Rather than configuring "raw"CLI commands interface- by-interface, you
develop a uniform setof traffic classes and QoS policies thatcan beapplied on interfaces.
fhe use ofthe MQC allowsthe separation of trafficclassification fromthe definition of QoS
policy. This enables easier initial QoS implementation andmaintenance as newtraffic classes
emerge and QoS policies for the network evolve.
>2010 Cisco Systems, Inc. Quality of Service
Implementing QoS with Cisco AutoQoS
Thissublopic explains CiscoAuloQoS. an easy-to-use automated tool to provision QoSacross
the entire network.
Implementing QoS with Auto<
WAN
AutoQoS VoIP supported both in
the LAN and WAN environments
AutoQoS Enterprise supported
on WAN interfaces
Routers can deploy Enterpnse
QoS policy treatment for voice,
video, and data traffic
Switches can deploy QoS policy
treatments for voice by a single
command
The [trust] option indicates that
the DSCP or CoS markings of a
packet are relied upon for
classification AutoQoS Enterprise:
JtO dia
jtc qos
ery aos [trust]
There arc these two versions of Cisco AutoQoS:
Cisco AutoQoS VoIP: in its initial release. Cisco AuloQoS Vol!1 providedbest-practice
QoSconfiguration for VoIP on bothCiscoswitches and routers. This was accomplished by
entering one global or interface command. Depending on the platform, the Cisco AuloQoS
macro would then generate commands into the recommended VoIP QoS configurations,
along with class maps and policy maps, and apply those to a router interface or switch port.
Cisco AutoQoS is available on both LANand WANCisco Catalyst switches and Cisco
IOS routers,
Cisco AutoQoS for the F.nterprise: Cisco AuloQoS for the Enterprise relics on NBAR to
gather statistics and detect 10 traffic types, resulting in the provisioning of class maps and
policy maps for these traffic tvpes. This feature deploys best-practice QoS policies lor
voice, video, and data traffic. Altogether. 10 traffic types are detected as traffic crosses the
WAN interfaces.
Cisco AutoQoS for the Enterprise, combined with the auto qos voip command, allows a novice
network administrator to administer complex, detailed QoS policies throughout the enterprise
network. Cisco AutoQoS for the Lnlerprise works only for Cisco IOS router platforms. The
VoIP feature for Cisco Catalyst switches docs not change.
Ihcre are some major differences between Cisco AutoQoS VoIP and Cisco AuloQoS for the
Lnlerprise. Cisco AutoQoS VoIP does not detect traffic types, nor docs it use NBAR. Cisco
AutoQoS VoIP only creates QoS policy to provide priority of voice traftic. Cisco AutoQoS for
the Enterprise, on the other hand, uses a discovery mechanism or traffic data collection process
that uses NBAR. The Cisco AutoQoS VoIP macros use the NBAR statistics to create QoS
policies.
Implementing Cisco Voice Communications and OoS (CVOICE) v8 0 )2010 Cisco Systems, Inc
Comparing QoS Implementation Methods
This subtopic compares the methods of implementing QoS policy in the enterprise^network.
Comparing QoS Implementation
Ease of use
Abilityto
fine-tune
Time to
deploy
Modularity
Poor
OK
Longest
Poor
AutoQoS AutoQoS
VoIP Enterprise
Easier Simple Simple
Very good Very good Very good
Average Shortest Shortest
Excellent Excellent Excellent
Cisco recommends the use of MQC and Cisco AutoQoS VoIP when deploying voice over the
LAN and Cisco AutoQoS for the Enterprise on router WAN interfaces.
While MQC is much easier to use than CLI, Cisco AutoQoS VoIP and Cisco AutoQoSjfo,"the
Enterprise can simplify the configuration of QoS. As aresult, you can accomplish the lastest
implementation with Cisco AutoQoS.
MQC offers excellent modularity and the ability to fine-tune complex networks. Cisco
AutoQoS offers the fastest way to implement QoS, but has limited fine-tuning capabth.es.
wZa Cisco AutoQoS configuration has been generated, you must use CLI commands to
fine-tune aCisco AutoQoS configuratton. if necessary. (On most networks, fine-tumng w.li not
benecessary for Cisco AutoQoS.)
Quality of Service
QoS Models
5-28
This topic describes three models of QoS implementation.
Three Models for Quality of Service
Best Effort: No QoS isapplied to packets
IntServ Applications signal to the network that they require
special QoS.
DiffServ: The network recognizes classes that require
special QoS.
The follow ing three models exist for implemenling QoS in anetwork:
- Best-effort: With the besl-ef.brt model. QoS is not applied to packets. If it is not important
when or how packets arrive, the best-effort model is appropriate.
IntServ: Integrated Services (In(Serv) can provide high QoS lo IP packets. EssenttalK
applications signal to the network that they will require special QoS for aperiod of time
and that banihv.dth ,s reserved. With IniScrv. packet delivery is guaranteed. However the
use ot InlSen can severely limit the scalability ofa network.
DiffServ: Differentiated services (DiffServ) provide the greatest scalability and flexibility
in implementing QoS in anetwork. Network devices recognize traffic classes and provide
dtiferent levels ot QoS to different traffic classes.
Best-Effort Model
This subtopic describes the best-effort model, which provides the easiest way toaddress QoS.
Best-Effort Model
- Internet initially based on a
best-effort packet delivery
service
The default mode for all traffic
No differentiation between
types of traffic
Like using standard mail
It willget there when it gets there.
TheInternet wasdesigned forbest-effort, no-guarantee delivery of packets. This behavior is
still predominant on the Internet today.
If QoS policies arenot implemented, traffic isforwarded using thebest-effort model. All
network packets are treated exactly thesamean emergency voice message is treated exactly
likea digital photograph that is attached to anemail. Without the implementation of QoS. the
network cannot tell the difference between packets and, as a result, cannot treat packets
preferentially.
When youdropa letter instandard postal mail, youare using a best-effort model. Yourletter
will betreated exactly thesame as every other letter; it will get there when it getsthere. With
the best-effort model, the letter may actually never arrive. Unless you have a separate
notification arrangement withthe letter recipient, youmay neverknow if the letter doesnot
arrive.
Quality of Service
IntServ Model
This subtopic describes the IntServ model, which provides themostcomplex way toaddress
QoS.
Some applications have special
bandwidth and delay
requirements
Requests specific kind of
service from the network
before sending data
IntServ guarantees a predictable
behavior of the network for these
applications
Uses RSVP to reserve network
resources
Guaranteed delivery: No other
traffic can use reserved
bandwidth
Like having your own private
courier plane
It will getthere by 1030 a.m
Some applications, such as high-resolution video, require consistent, dedicated bandwidth to
prov idesufficient quality for \ iewers. IntServ was introduced to guarantee predictable network
behavior for theseapplications. Because IntServ reserves bandwidth throughout a network, no
other traffic can use the reserved bandwidth. Bandwidth that is unused, but reserved, is wasted.
IntServ is similar to a concept known as ""hard QoS." With hard QoS, traffic characteristics
suchas bandwidth, delay, andpacket-loss rales are guaranteed end-to-end. Thisguarantee
ensures both predictable and guaranteed service levels for mission-critical applications. There
will be no impact on traffic when guarantees are made, regardless of additional network traffic.
Hard QoS is accomplished by negotiatingspecific QoS requirements upon establishment of a
connection and bv using Call Admission Control (CAC) to ensure that no new traffic will
violate the guarantee.
Using IntServ is likehav inga private courierairplane or truckthat is dedicated to thedelivery
of your traffic. This model ensures quality and delivery, is expensive, and is not scalable.
IntServ is a multiple-sen.ice model that can accommodate multiple QoS requirements. IntServ
inherits the connection-orientedapproach fromtelephony networkdesign. Every individual
communicationmust explicitly specify its traffic descriptor and requested resources to the
network, "fhe edge router performs admission control to ensure that available resources are
sufficient in the network. The IntServ standard assumes that routers along a path set and
maintain the state for each individual communication.
The role of Resource Reservation Protocol (RSVP) is to provide resource admission control for
VoIP networks. If resources are available. RSVP accepts a reservation and installs a traffic
classifier in the QoS forwardingpath. The traffic classifier tells the QoS forwardingpath how
to classify packets from a particular flowand what forwardingtreatment to provide.
DiffServ Model A ,,.
way toprovide QoS.
Network traffic identified byclass
Network QoS policy enforces
differentiated treatment oftraffic
classes
You choose level ofservice for
each traffic class
Like using a package delivery
service
Do you want overnight delivery9
Do you want two-day air delivery?
Do you want three-toseven-day
grounddeiivery?
DiffServ was designed to overcome the limitations of both the best-effort and IntServ'models.
Difflcrv "n provfde an '"almost guaranteed" QoS, while still being cost-effective and scalable.
DiffServ is similar to aconcept known as "soft QoS." With soft QoS, QoS ^ntsms are
^T^lucllpro^to implementing QoS than hard QoS, because many (hundreds or
rtenualv thousands) of app ieations can be mapped into asmall set ol classes upon which
sS: eL of OoS tehaviS are implemented. Although QoS mechanisms ,n th.s approach are
entrced a*:I applieôn ahop-by-hop basis, uniformly applying global meanmg to each traflic
class provides both flexibility and scalability.
With DiffServ network traffic is divided into classes that are based on business requirements,
fac oSdass- can then be assigned adifferent level of service As the P-^jo*a
network each ofthe network devices identifies the packet class and services the packet
ac3h.p to Uiat class. You can choose many levels of service with DiffServ. For example,
ISfrom IP phones is usually given preferential treatment over all other appl-tmn
traffic. Lmail is generally given best-effort service. Nonbusiness traffic can other be given very
poor service orblocked entirely.
DifBerv v.rks like apackage delivery servie, You request <-d Pay for) alevel of service
v.hen vou send your package. Throughout the package network, the level or service is
^cognized and your package is given either preferential or normal scrv.ee, depend.ng on what
yourequested.
Quality of Service
6-32
QoS Model Evaluation
Thissubtopicevaluates the three QoS models by comparing their benefits and drawbacks
1
DiffServ
IntServ
Noabsolute serviceguarantee
Complex mechanisms
Continuous signaling because of
stateful architecture
* Flow-based approach notscalable
to targe implementations such as
ISP networks or the Internet
Best effort No service guarantee
No service differentiation
Highlyscalable
Many lewisofquality possible
* Explicit resource admission
control (end to end)
Per-requestpdicy admission
contrrJ
* Signaling ofdynamic port
numbers (forexample, H.323)
* Highlyscalable
* No special mechanisms
required
DiffServ has these kev benefits:
It is highly scalable.
It prov ides many different levels ofqualitv.
DiffServ also has thesedrawbacks:
No absolute guarantee ofserv ice quality can be made.
It requires aset of complex mechanisms to work in concert throughout the network.
The main benefits ofIntServ and RSVP are as follows:
' ,RSth Psi.gn.als.PS ûcsts ^ ^^vidual flow. The network can then provide guarantees
othese individual lows fhe problem with this is that IntServ does not scale to'argf
networks because of the large number ofconcurrent RSVP Hows.
' aRS7 ffmiS ne,tWOrk.devices of ilow Parameters (IP addresses and port numbers) Some
apphcaaons use vnamtc port numbers, which can be difficult for network devices to
recognize. NBAR is amechanism that has been introduced to supplement RSVP for
applications that use dvnamic port numbers but do not use RSVP.
' '"svT suPPrtifadmis^" control, which allows anetwork lo reject (or downgrade) new
RSVP session,,f one of the interlaces in the path has reached the limit (that is all
reservablc bandwidth is booked).
The main drawbacks of IntServ and RSVP are as follows:
fhere is continuous signaling because ofthe stateful RSVP operation.
' ,nS,h P^"If S?,ab'C l '"^ netWOrkS Wherc per"flow gUarantees *ouW havc to be made
to thousands ot concurrent RSVP flows.
2010 Cisco Systems, tne
The best-effort model has these significant benefits:
Thebest-effort model has nearly unlimited scalability, fhe onlywayto reach scalability
limits is to reachbandwidth limits, in whichcase all trafficbecomesequallydelayed.
You donotneed toemploy special QoS mechanisms to use the best-effort model. It isthe
easiest and quickest model to deploy.
The best-effort model also has these drawbacks:
Notliing isguaranteed. Packets will arrive whenever they can, inany order possible, if they
arrive at all.
Packets arenot given preferential treatment. Critical datais treated thesame as email.
) 2010 Cisco Systems. Inc. Quality of Service 6-33
Summary
This topic summari/es the key points that were discussed in this lesson.
Summary
The most critical QoS issues include lack of bandwidth, end-to-end
delay, jitter, and packet loss.
QoS policydefines the specific levels of quality of service assigned
to different classes of network traffic.
QoS is implemented in Cisco Unified Communications networks by
identifying traffic, dividing trafficinto classes, and assigning QoS
policies to the classes
One-way QoS requirements for VoIP traffic define maximum
latency (150 ms), jitter (30 ms), loss (1 percent), and guaranteed
priority bandwidth per call (17-106 kb/s).
QoS can be implemented using CLI. MQC, AutoQoS, or QPM.
The three QoS models are best effort, IntServ, and DiffServ.
Lesson 2\
Understanding QoS
Mechanisms and Models
Overview
Differentiated services (DiffServ) is amultiple-service model for implementing quality of
service (QoS) in the network. With DiffServ, the network tries to deliver aparticular kind of
en i" ihat is based on the QoS specified by each packet. This ^^.^"^
different wavs. such as using the differentiated serv,ces code point (DSCP) in IP packets o
source and destination addresses. The network uses the QoS specfication ***
classifv. shape, and police traffic and to perform intelligent queuing. This leswn focuses on the
DiffServ model and explains the mechanisms that are used to implement DiftServ.
Objectives
Upon completing this lesson, you will be able to describe the characteristics and mechanisms of
the Difl^n'model and contrast the DiffServ model to other models. This abthty includes betng
able to meet these objectives:
Explain the purpose and function ofDiffServ
. Describe the basic format and the purpose ofthe DSCP field in the IP header and contrast it
tothetraditionally used IPprecedence format
List the different per-hop behaviors that are used in DSCP
. Describe the interoperability between DSCP-based and IP precedence-based devices
Explain the key mechanisms of DiffServ to implement QoS in an IP network
Describe Cisco QoS baseline model for enterprise
DiffServ Model
6-36
This topic describes the DiffServ model and tcrminolom
' S'SSqoS JS"de POmt (DSCP>: AValUe m** 'Pheaded
' (BoneaS!,cdass)a,e (BA'" A^^'^f P3CketS Wi'h 'he Same DSCP
'2Z7*tTT(PHBi" ThB fowardinS behavior (QoS treatment)
applied at a BA by a node
BA 1
1 \fcice
| DSCP 46
Vq*c
DSCP 46
Vorce
DSCP 46
BA2
IfTPftsmAtoBl
1 OSCP22 |
FTP tram AtoB
DSCP 22
PtPffumAloB
'QSCP22
Apply PHB X lo
Appfy PHB C lo 6A2
Apply PHB V to I
Apply PHB BIOBA2
Apply PHB Zto BA 1|
Apply PHBAtoBA2
The discussion about the DiffServ model uses three basic terms to describe DiffServ
operations, as follows:
Behavior aggregate (BA): ABA is acollection of packets with the same DSCP value
crossing alink ,n aparticular direction. Packets from multiple applications and sources can
belong to the same BA. In Csco IOS Software, classification of packets into BAs can be
done bv using Modular QoS command-line interface (CLI). or MQC, class maps.
' nSv?: A*al,U? 'Ithc 'P hcadcr that iS used t0 sc,ect aQs lreillm1 for apacket. In the
UtfServ model, classification and QoS revolve around the DSCP.
Per-hop bcha. ior (PHB): APHB is at. externally observable forwarding behavior (or QoS
treatment) applied at aDiffServ compliant node to aDiffServ BA. The term PHB refers to'
the packet scheduling, queuing, policing, or shaping behavior of anode on any given
packet belonging to aBA. The DiffServ mode! itselfdoes not speeifv how PHBs musl be
'tr^V^ *anet> of tcchni<-ues y be used to affect the desired traffic conditioning
and IMB. In Cisco IOS Software, you can configure PHBs bv using MQC policv maps
Implementing Cisco Voice Communications and QoS (CVOICE) v8.D
DiffServ Model
This subtopic describes the approach to QoS in the DiffServ model.
DiffServ Model
QoS behaviors applied to traffic classes on a per-hop basis.
Com plex traffic classification and conditioning performed at network edge:
- Network traffic is categorized into BAs
Each packet belonging to a BA is marked with a DSCP value.
Network devices in the core use the DSCP value to select a per-hop
behavior for the packet
Classify ana
mark witti
DSCP at
network edge
Match DSCP Match DSCP Match DSCP
anfl select per- and selecl per- and select per-
hop-betsawor fiop -behavior, hop-behavior.
The DiffServarchitecture is based on a simple model in which traffic entering a network is
classified and possibly conditioned at the boundaries ofthe network. The traffic class is then
identified with a DSCPor bit marking in the IP header, 'fhe primary advantage of DiffServis
scalability.
DSCP values are used to mark packets to indicate a desired PHB. Within the core ofthe
network, packets are fonvarded according to the PHB that is associated with the DSCP.
One of the primary principles of DiffServ is that packets should be marked as close to the edge
ofthe network as possible. It is often a difficult and time-consuming task to determine the
trafficclass for a datapacket, so the datashouldbe classified as fewtimes as possible. By
markingthe trafficat the network edge, corenetworkdevicesand other devicesalongthe
forwarding path will be able to quickly determine the properQoStreatmentto applyto a given
traffic flow.
DSCPsupersedes IP precedence, a 3-bit field in the type of service (ToS) byte ofthe IP header
that was originally used to classify and prioritize types of traffic. However, DSCP maintains
interoperability with devices that use IP precedence.
>2010Cisco Systems. Inc.
DSCP Encoding
This topic describes the use ofthe ToS byte for DSCP.
DSCP Encoding
DS field The IPv4 header ToS octet or the IPv6 traffic class octet, when
interpreted in conformance with the definition given in RFC 2474
DSCP The first six bits of the DS field, used to select a PHB (forwarding
and queuingmethod)
ToS
Byte
Flags Oftse! TTL 1 Prao
IPv4 Packet Header
IPv4 IP Precedence
' DS Field
DiffServ uses the differentiated services field (DS field) in tlie IP header to mark packets
according totheirclassification into BAs. The DS field occupies thesame 8 bitsofthe IP
header that vv ere previously used for the ToS bv te.
fhe following three Internet Fngineering Task Force (lblT) standards describe thepurpose of
the 8 bits ofthe DS field:
RFC 791 includes specification ofthe ToS field, where thethree low-order bitsare used for
IPprecedence. The next 3 bitsare used for delay, throughput, reliability, andcost.
RFC 1812 modifies the meaning ofthe ToS field by removing the meaningfrom the live
high-order bits(those bitsshould all be0). Thisgained widespread use andbecame known
as the original IP precedence.
RFC 2474 replaces the ToS field wilh the DS field, where the sixlow-order bits areused
for the DSCP. Fhe remaining 2 bitsare used for explicit congestion notification. RFC 3260
{Sew Terminology andClarificationsfor DiffServ) updates RFC 2474 andprovides
temiinologv clarifications.
Fach DSCPvalue identifies a BA. Fach BA is assigned a PHB. Each PHB is implemented
using the appropriate QoS mechanisms.
IPversion 6 (IPv6) alsoprovides support for QoS marking viaa field inthe IPv6 header.
Similar to ihe ToS (or DS) field in the IPv4 header, the TrafficClass field(8 bits) is available
for use by originating nodes and forwarding routers toidentify and distinguish between
different classes or priorities of IPv6 packets, fhe Traffic Class field can beused tosetspecific
precedence or DSCP values. which arc used the same way that they are used in IPv4.
DiffServ PHBs
This topic describes the concept of aPHB.
Per-Hop Behaviors
DSCP selects PHB throughout the network.
. DefaultPHB:(FIFO, taildrop)
EF: Expedited Forwarding
- AF: Assured Forwarding
Class-Selector: (IP precedence) PHB
0 1 2 3 *__L_
K^Bp^BssHsBISdscp
000 = Default
101 =Expedited Forwarding
001, 010, m1. or100 =Assured Forwarding
000 = Class Selector
fhe IETF standards define the following PHBs:
Default PHB: Used for best-effort service (bits 0to 2of DSCP =000)
. Expedited Forwarding (EF) PHB: Used for low-delay service (bits 0to 2of DSCP =
l()l)
Assured Forwarding (AF) PHB: Used for guaranteed bandwidth service (bits 0to 2of
DSCP = 001. 010. 011, or 100)
- Class-Selector PHB: Used for backward compatibility withnon-DiffServ-compliant
devices (RFC 1812-compliant devices [bits 3to 5of DSCP - 000])
Quality of Service
6-40
Expedited Forwarding PHB
This subtopic explains the Expedited Forwarding (EF) PHB.
EF PHB
Ensuresa minimum departurerate
Guarantees bandwidth (the class is guaranteed an amount
of bandwidth with prioritized forwarding)
Polices bandwidth (excess traffic is dropped)
DSCP value 101110,
Looks like IP precedence 5to non-DiffServ devices
Bits 0to 2 101 =5(same bits as tor IP precedence)
Bits 3to4-11 (same bits as drop probability, fixed value in EF PHB1
Bit 5 JustO
No Drop
Probability
DSCP
fhe FF PHB is identified based on the following;
The FF PHB ensures aminimum departure rate. The FF PHB provides the lowest possible
delay to delay-sensitive applications. possmie
- The FF PHB guarantees bandwidlh. The FFPUB prevents starvation ofthe application if
there are multiple applications using FF PHB.
' I?CthF PHI?.P!ices ban!,widlh whcn congestion occurs. The EF PHB prevents starvation
ol other applications or classes that are not using this PHB.
Packets requiring FF should be marked with DSCP binary value I0l 110 (46 or 0x2F).
Non-DiH^rv-compliant devices uill regard EF DSCP value 101110as IP precedence 5(101)
Ih.s precedence ,s the highest user-definable 1P precedence and is typically used for delay -
sensitive traffic (such as VoIP). Ihe three low-order bits ofthe FF DSCP value are 101 "hich
matches IP precedence 5and allows backward compatibility
Implementing Cisco Voice Communicalions and QoS (CVOICE) v8 0
This subtopic explains the Assured Forwarding (AF) PHB.
AFPHB:
~- Guarantees bandwidth
- Allows access toextra bandwidth, if available
Four standard classes (af1. af2, af3, and af4)
DSCPvalue range: aaaddO
- Whereaaa is a binary valueof the class
- Where dd is drop probability
iMdscp
aaa
The AF PHB is identified based onthe following:
The AF PHB guarantees acertain amount of bandwidth lo an AF class.
The AF PHB allows access to extra bandwidth, ifavailable.
. Packets requiring AF PHB should be marked with DSCP value aaaddO, where aaa is the
number ofthe class and dd isthe drop probability.
There are four standard defined AF classes. Each class should be treated independently and
should have allocated bandwidth that is based on the QoS policy.
6-42
Assured Forwarding PHI
Each AFclass uses three DSCPvalues.
Each AF class is independently forwarded with its guaranteed bandwidth.
Congestion avoidance is used within each class (drop probability)
dscp =afh
Class
Value
AF1 001 dd 0
AF2 010 dd 0
AF3 011 dd 0
AF4 100 dd 0
Drop
Probability
(dd)
Value
AF
Value
Low
01 AF11
Medium 10
AF12
High 11
AF13
Fad, AF class is assigned an IP precedence and has three drop probabilities: low. medium, and
AF.vt represents an AF' PHB. where .v corresponds to the IP precedence value (onlv IP
precedences Ito 4are used for AF classes), and ycorresponds to the drop preference value (I
2. or 3).
Implementing Cisco Voice Communicalions and QoS (CVOICE) vS.i
>2010Cisco Systems.
Assured Forwarding PHB (Cont.)
* AF PHB does not followthe "bigger-is-betler" logic
- For example: AF11 (decimal 10) and AF13 (decimal 14)
Same queuing class, but AF11 "better" due to lower drop
Probability
Low drop
probatriity
AF11
001010
Decimal 10
AF21
010010
Decimal: 18
AF31
011010
Decimal: 28
AF41
100010
Decimal: 34
Mediixn drop
probabiEty
AF12
001100
Decimal 12
AF22
010100
Decimal: 20
AF32
011100
Decimal: 28
AF42
100100
Decimal: 36
High drop
prababiity
AF13
001110
Decrnal 14
AF23
010110
Decimal: 22
AF33
011110
Decimal: 30
AF43
100110
Decimal: 38
Interestingly, the AFPHBvalues do not necessarily follow the "bigger-is-bettcr" logicthat has
been usedwith IPprecedence marking, in whicha packetwith a higher IPprecedence received
preferential treatment over a packet with lower IP precedence.
The tablein the figure lists the binaryand decimal valuesfor the 4 AF PHBs in three drop
probability combinations. Within each class, a higher DSCPvaluesignifiesa higherdrop
probability, and therefore a less-preferential treatment.
This topic describes the DiffServ class selector.
Class-Selector xxxOOO DSCP
* Backward compatibility with IP precedence
Maps IP precedence to DSCP
- Differentiates probability of timely forwarding
- (xyzOOO) >= (abcOOO) if xyz > abc
- Ifa packet has DSCP 011000, it has a greater probability
of timely forwarding than a packet with 001000.
IP Precedence
IPv4 IP Precedence
Class
Selector
fhe meaning ofthe 8 bits in the DS field ofthe IPpacket haschanged over time tomeet the
expanding requirements of IPnetworks. Themost common traditional use is provided by the
three low-order IP precedence bits.
TheClass-Selector PUR provides backward compatibility tor DSCPwith IPprecedence. The
next 3 bits ofthe DSCP (bits 3 to 5). set to I), identify a Class-Selector PHB. The Class-Selector
PHB is defined as the probability of timelj forwarding andis compliant with RFC 1812. which
simply prioritizes packets according to theprecedence value. Packets with higher IPprecedence
should general!} be forwarded inlesstime than packets with lower IPprecedence.
DiffServ QoS Mechanisms
This topic descnbes the mechanisms that are implemented when deploying aDiffServ model.
QoS Mechanisms
Classification: Each class-oriented QoS mechanism has to
support some type of classification.
- Marking: Used to mark packets based on classification,
metering, or both.
Congestion management: Each interface must have a
queuing mechanism to prioritize transmission of packets.
. Congestion avoidance: Used to drop packets early to avoid
congestion later in the network.
Policing and shaping: Used to enforce a rate limit based on
the metering (excess traffic is either dropped, marked, or
delayed).
Link efficiency: Used to improve bandwidth efficiency through
compression, link fragmentation, and interleaving.
The main categories of tools that arc used to implement QoS in anetwork include the
following:
. Classification and marking: The identifying and splitting of traffic into different classes
and the marking of traffic according to behavior and business policies.
Congestion management:
based on markings.
Congestion avoidance: Discards specific packets that are based on markings to avoid
network congestion.
Policing and shaping: Traffic conditioning mechanisms that police traffic by dropping
misbehaving traffic to maintain network integrity. These mechanisms also shape traffic to
control bursts byqueuing traffic.
Link efficiency: One type of link efficiency technology is packet header compression.
which improves the bandwidth efficiency of alink. Another technology is link _
fragmentation and interleaving (LFI), which can decrease the "jitter ot vo.ce transmission
by reducing voice packet delay.
'fhe prioritization, protection, and isolation of traffic that is
Quality of Service
6-45
Classification
6-46
This subtopic explains the process ofpacket classification.
Classification is the identifying and splitting of traffic into
different classes.
Traffic can be classed by various means, including the DSCP.
Modular QoS CLI allows classification to be implemented as
a building block.
Classification is the identifying and splitting of traffic into different classes. In aQoS-enabled
network, all traffic ,s classified at the input interface of every QoS-aware device Packet
'"assittcatjon canbe based onmany factors, such as these:
DSCP
IP precedence
Source address
Destination address
The concept oft, is the key for deploying QoS. When an end device (such as aworkstation
or aCisco Unified IP phone) marks apacket with class of service (CoS) or DSCP aswitch or
router has the option of accepting or ignoring those values. If the switch or router chooses to
accept ihM alucs. the switch or router trusts the end device. If the switch or router trusts the end
device. ,t does not need to do any reclassification of packets coming from that interface If the
switch or router does not trust the interface, it must perform areclassification to determine the
appropriate QoS value tor the packets coming from that interface. Switches and routers are
generally set to not trust end devices and must specificallv he configured to trust packets
coming from an interface. Classification tools include Network-Based Application Recognition
(NBAR). policy-based routing (PBR). and classification and marking using MQC
Note
The tools for classification and other QoS mechanisms are covered in detail in the following
lessons.
Marking
This subtopic describes the purpose of marking and identifies where marking is commonly
implemented in a network.
Marking
Marking each packet as a member of a network class
- Allows instant recognition throughout the network
- Also called coloring
Marking, also known as coloring, involves marking each packet as a member of a network class
so that devices throughout the rest ofthe network can quickly recognize the packet class.
Marking is performed as close to the network edge as possible and is typically done using
MQC.
QoS mechanisms set bits in the DSCP or IP precedence fields of each IP packet according to
the class that the packet is in. Other fields can also be marked to aid in the identification of a
packet class.
Other QoS mechanisms use these bits to determine how to treat the packets when they arrive. If
the packets are marked as high-priority voice packets, the packets will generally never be
droppedby congestion avoidance mechanisms and will be given immediate preference by
congestion management queuing mechanisms. On the other hand, if the packets are marked as
low-priority file transfer packets, they will be dropped when congestion occurs and will
generally be moved to the end ofthe congestion management queues.
Quality of Service
This subtopic explains the concept of congestion management.
Evaluates packet marking to determine in which queue to place
packets
Sophisticated queuing technologies ensure that time-sensitive
packets (voice) are prioritized, such as WFQ and LLQ
Voice
d Mission-Cnlicai
Transactional
Voice Queue (First Out)
Mission-Critical Queue (40% bandwidth)
Transactional Queue (20% bandwidlh)
Congestion management mechanisms (queuing algorithms) use the marking on each packetto
determine in which queueto placepackets. Different queuesare givendifferent treatment by
the queuingalgorithm that is basedon the class of packetsin the queue. Cicnerally. queues wilh
high-priority packets receive preferential treatment.
Congestion management is implemented on all output interfaces in a QoS-enabled network by
usingqueuing mechanisms to managethe outflowof traffic, liachqueuing algorithm was
designed to solvea specific network trafficproblem and has a particular effect on network
performance.
The CiscoIOS Software for congestion management or queuingincludes these queuing
methods:
FIFO, priority queuing (PQ). custom queuing (CQ)
Weighted fair queuing (WFQ)
Class-based weighted fair queuing (CBWFQ)
Low latency queuing (LLQ)
LLQis currently the preferred queuing method. LLQ is a hybrid (PQ and CBWFQ) queuing
method that was developed to specifically meet the requirements of real-time traffic, suchas
\oiee.
Congestion Avoidance
This subtopic describes congestion avoidance and identifies where congestion avoidance is
commonly implemented in anetwork.
Congestion Avoidance
Random packet drop from selected queues when previously defined
limits are reached
Helps prevent bottlenecks downstream in the network
. Can be implemented without drop using explicit congestion notification
(ECN)
- Two remaining bits inToSfield
Voice Queue (First Out)
Mission-Critical Queue (40%bandwidth)
TransactionalQueue (20%bandwidth)
Congestion-avoidance mechanisms monitor network traffic loads in an effort to anticipate and
avoid congestion at common network bottlenecks. Congestion avoidance is ach.eved through
packet dropping.
Congestion-avoidance mechanisms are typically implemented on output interfaces where a
high-speed link or set of links feeds into alower-speed link (such as aLAN feeding into a
slower WAN link). This ensures that the WAN is not instantly congested by LAN traffic.
Weighted random early detection (WRED) is aCisco primary congestion-avoidance technique.
WRED increases the probability that congestion is avoided by dropping low-pnonty packets
rather than dropping high-priority packets.
WRED is not recommended for voice queues. Anetwork should not be designed to drop voice
packets.
Policing
This subtopic explains the concept of traffic policing.
Policing
Dropping or marking of packets when a predefined limit is
reached
Protection of other traffic classes to ensure that thev do not
starve
Policing is used to condition traffic before transmitting traffic to anetwork or receiving traffic
from a network.
Policing is the ability to control bursts and traffic to ensure that certain types of traffic get
certain types of bandwidth.
Policing drops or marks packets when predefined limits are reached. Policing mechanisms can
be set to first drop traffic classes that have lower QoS priority markings.
Policing mechanisms can be used at either input or output interfaces. These mechanisms are
typically used to control the flow into anetwork device from ahigh-speed link bv droppin-
excess low-pnonty packets. Agood example would be the use ofpolicing bv aservice provider
to throttle ahigh-speed inflow from acustomer that was in excess ofthe service agreement In
a ICPenv ironment. this policing would cause the sender lo slow its packet transmission.
Tools include class-based policing and committed access rate (CAR).
Shaping
This subtopic explains traffic shaping.
Shaping
- Queuing of packets when a predefined limit is reached
. Forwarding of temporary bursts that would have to dropped
as above-the-threshold iflong-termaverage limit not
exceeded
UNDER LIMITAGAIN
(Use Buffered Packets)
Shaping helps smooth out speed mismatches in the network and limits transmission rates.
Shaping mechanisms are used on output interfaces. These mechanisms are typically used to
limit the flow from ahigh-speed link to alower-speed link to ensure that the lower-speed link
does not become overrun with traffic. Shaping could also be used to manage the flow oftraffic
at apoint in the network where multiple flows are aggregated. Service providers use shaping to
manage the flow of traffic to and from customers to ensure that the flows conform to service
agreements between the customer and provider.
Cisco QoS software solutions include two traffic-shaping tools to manage traffic and
congestion on the network: generic traffic shaping and Frame Relay traffic shaping (FRTS).
Quality of Service
Compression
rhis subtopic explains how lo reduce bandwidth consumption by compression.
Compression
' Reducing the overhead associated with voice transport
Mapping of IP/UDP/RTPheader to 2-octet index
Optional 2-octet checksum
Bandwidth-saving mechanism tosupport large amount oftraffic
over a slow link
ReduceVoice Headerto 2or 4 Bytes
2 or 4 Bytes 20 Bytes
Compression is one ofthe Cisco IOS link-efficiency mechanisms that work in conjunction with
queuing and traffic shaping to manage existing bandwidth more efficiently and predictably.
Two types of compressionare available:
Compression of the pay load of I.ay er 2 frames. One of two algorithms. Stacker or
Predictor, can beconfigured for this type of compression.
Compressed Real-Time Transport Protocol (cRTP), as shown in the figure, maps the three
headers. IP. User Datagram Protocol (UDP). and Real-Time Transport Protocol (RTP).
with acombined 40 by tes. to 2or 4bytes, depending on whether the cyclic redundancy
check (CRC) is transmitted, 'fhis compression can dramatically improve the performance
of a link.
Compression should only be used on slow WAN links because its drawback isthe consumption
of computational resources on a hop-by-hop basis.
Link Fragmentation and Interleaving
This subtopic explains LFI.
Link Fragmentation and interleaving (LFI)
Breaks long data packets apart
Interleaves delay-sensitive packets so that theyare timely
forwarded rather than being clogged behind largedata
packets
Reduces jitterfor delay-sensitive traffic (voice)
packets.
Interactive traffic, such as VoIP, issusceptible to increased latency and jitter when the network
processes large packets, such as LAN-to-LAN FTP Telnet transfers traversing a WAN link,
'fhis susceptibility increases asthe traffic is queued onslower links.
LFI can reduce delay and jitter on slower-speed links by breaking up large datagrams and
interleaving low-delay traffic packets with the resulting smaller packets.
LFI is used on slow WAN links toensure minimal delay for voice and video traffic.
Applying QoS to Input and Output Interfaces
This subtopic describes which mechanisms can be applied toinput and which tooutput
interfaces.
(As close to the
source as possible)
(Coming from a
higher-speed ink or
aggregation)
Classify
Policing
Congestion
Management
Mari(
Congestion
Avoidance
Shaping
- >
Policing
Compression
Fragmentation
and Interleaving
(Always)
(High-speed to
low-speed links or
aggregation points)
(Low-speed
WAN links)
In aQoS-enabled network, classification is performed on every input interface. Marking should
be perfomied as close to the network edge as possiblein the originating network device, if
possible. Devices further from the edge ofthe network, such as routers and switches, can be
configured to trust or ignore the marking set by the edge devices. ACisco Unified IP phone, for
example, will not trust ihe markings of an attached PC. while switches are typically configured
totrustthemarkings of attached Cisco Unified IPphones.
It only makes sense to use congestion management, congestion avoidance, and traffic-shaping
mechanisms on output interfaces. Ihese mechanisms help maintain smooth operation of links
by controlling how much and which type oftraffic is allowed on alink.
Congestion a% oidanee is npicallv employed on an output interface where there is achance that
ahigh-speed link or aggregation of links feeds into aslower link (such as aLAN feeding into a
WAN).
Policing and shaping are t\ picallv employed on output interfaces to control Ihe flow of traffic
from ahigh-speed link to lower-speed links. Policing is also employed on input interlaces to
control the flow into anetwork device from ahigh-speed link by dropping excess low-priority
packets.
Both compression and LFI are typically used on slower-speed WAN links between sites lo
improve bandwidth efficiency.
6-54 Implementing Cisco Voice Communications and OoS (CVOICE) v8 0
"mm
"mm
-mw
Cisco QoS Baseline Model
This topic describes Cisco QoS baseline model and its variants.
Cisco Baseline Classification
Description
Traflic Class
Voice
Videoconferencing Interactive video date traffic
Streaming Video Streaming media traffic
Missiof^CrtacalData AppBcatJomwith offlcat importance t
Call Signaling Call signaling and control traffic
Network Management Network management traffic
Bulk Data
Default clnss all noncriiical traffic_
Although there are several sources of information that can be used as guidelines for
d ^tiifg aQo? policy, none of mem can determtne exactly what ,s proper for aspecific
SSll*presents its own unique challenges and administrative pohe.es.
The Cisco baseline classification model provides one ofthe possible classification approaches.
I con sts of 11 traffic classes that are typically found in enterprise networks The 11 traffic
asTesie described in the table and provide enough granularity for amajority of
nVirions The model can grow or shrink based on enterprise requirements. It should
SZ^T^lLitli betw.cn traffic categories with the goal of easy manageability.
Quality of Service
6-55
6-56
Cisco Baseline Marking
1
Cisco Baseline Markim
Application
Voice
5
VideoConferencing
4
Streaming Video 4
Mission-Critical Data
3
Call Signaling
3
Transactional Data 2
Network Management
2
Bult Data
1
L3 Classification
PHB
AF41
CS4
AF31
CS3
&$
CS2
AF11
46
34
32
26
24
18
16
10
S'^ * * *e '' <* categories that are
OoS must be implemented consistently across the entire network. It is nol so important whether
Call Signaling ,s marked as DSCP 34 or 26. but rather lhat DSCP 34 and 26 a t aed a
DSCP 3,s' tre7ra- *T"^*< <** ^ "is ^ ''"POU-nt ^dlt^ ed
DSC P34 ,s reatedconsistently across the ne.work. If data travels over even asmall portion of
anetwork where different policies are applied (or no policies are applied, the ent ire OoS
Sicd^
mS'bv Ci ^rN' J3" Signa'ing tR,ffiC " AF 31: CaN Signaling traffic ori8inav
marked bv CtoIP lelephony equipment to DSCP AF31. The Cisco QoS baseline changed the
marking recommendation for call signaling traffic to DSCP CS3 because Class SdecSe
points, as defined ,n RFC 2474. are not subject to markdown or aggressive d^ppfng
Cisco Baseline Mechanisms

This subtopic describes the recommended actions that are used for the traffic classes that are
specified in the Cisco baseline model.
Cisco Basel ine Mechanisms
AppflCalkin j. RecoataendadAction
Rout-no Rate based queuing * RED
Voice CAC+prt0ftyqjeJn3
VldeoConferencing CAC +fate4jaTOd4Lts..v,; .".:*:-.
Streaming Video CAC + rate-based queuing + WRED
Msaion-CrHicatData Rate-basedqueuing * WRED
Call Signaling Rate-based queuing ^ RCD
S?^^.^^??W?"nflW3ia8^WWRffl isaiHHHKm'"':.'""
Network Management Rate-based queung + RED
BukData Rate-based queuing* WRED
Besi Eift on I Ljandv/idlli guarantee uate-tiased queuing +RED)
The table in the figure lists the QoS mechanisms that are recommended for each ofthe 11
traffic classes ofthe Cisco baseline model.
Call Admission Control (CAC) ensures that only a defined number of simultaneous calls are
admitted into the VoIP network.
Priority (LLQ) and rale-based queuing (CBWFQ) are discussed in detail in a later lesson.
Random Early Detection (RED) and WRED selectively drop packets when bulfers are filling
up. This random drop is used to avoid congestion and is most effective for TCP flows, which
reduce their transmit window size.
Quality of Service
Expansion and Reduction of Class Model
fhis subtopic describes how the class model is expanded and reduced based on the enterprise
requirements.
3-Class Model 5-Class Model 8-Class Model Baseline Model
us
Model selection based on enterprise requirements
The figure shows other common classification models, based on 3. 5, or 8 trafilc classes,
respectively. The expansion and reduction ofthe model causes some ofthe classes to be split or
aggregated into more granular or more generic traffic categories, fhe selection ofthe most
suitable model depends on enterprise needs.
6-58 Implementing CiscoVoice Communicationsand OoS (CVOICE) v8 0
Summary
This topic summarizes the key points that were discussed inthis lesson.
Summary
In the DiffServ model, QoS behaviors are applied to traffic classes
on a per-hop basis.
- DSCP uses the first six bits ofthe IPv4 ToS field or the IPv6 traffic
class field.
Aper-hopbehavior is forwarding behavior applied at a node to a
DiffServ behavior aggregate and can be classifiedintofour types:
Default, Expedited Forwarding, Assured Forwarding, and Class-
Selector.
The class-selector code point provides backward compatibility with
IP precedence.
* QoS mechanisms include classification, marking, congestion
management and avoidance, policing, shaping, compression,
fragmentation, and interleaving.
Cisco QoS baseline model identifies 11 traffic classes that typically
coverthe needs of every enterprise.
2010 Cisco Systems. Inc Quality of Service 6-59
Lesson 3
Explaining Classification,
Marking, and Link Efficiency
Mechanisms
Overview
"fhe Modular quality of service (QoS) command-line interface (CI.I), or MQC, provides a
modular approach to the configuration of QoS mechanisms. MQC allows network
administrators to introduce new QoS mechanisms and reuse available classification options.
MQC offers a scalable method to provide different levels of treatment to specific classes of
traffic. Before any QoS applications or mechanisms can be applied, traffic must be identified
and sorted into different classes. QoS is applied to these different traffic classes. Network
devices use classification to identify traffic as belonging to a specific class. After network
traffic is sorted, marking can be used to color (tag) individual packets so that other network
devices can apply QoS features uniformly to those packets as they travel through the network.
VoIP traffic is susceptible to latency when large packets, such as bulk FTP packets, traverse
WAN links. Packet delay is especially significant when large packets are queued on slower
links (less than or equal to 768 kb/s). To solve delay problems on slow bandwidth links, a
method for fragmenting larger frames and then queuing smaller frames between fragments of
the larger frames is required. To meet this requirement. Cisco IOS Software supports Multilink
PPP (MLP) link fragmentation and interleaving (ITT) and Frame Relay Fragmentation
Implementation Agreement (FRF. 12). In addition, complementary tools, such as header and
payload compression techniques, can be deployed to reduce the size of frames that are sent over
WAN links.
This lesson outlines how to implement QoS policies using MQC, and introduces the concepts
of classification and marking. It explains the different markers that are available at the data-link
and network layers, and identifies where classification and marking should be used in a
network. The lesson also describes different approaches for improving the efficiency of WAN
links.
Objectives
Uponcompleting this lesson, you will be able to describe the operation and configurationofthe
QoS classification and marking mechanisms, further, you will be able to explain trust
boundaries, marking based on class of service (CoS). differentiatedservices code point
(DSCP). and IP precedence, and mapping between Layer 2 and Layer 3 markers, "fhis ability
includes being able to meet these objectives:
Describe the three steps that are involved in implementing a QoS policy using MQC. and
the differences between class maps, policy maps, and service policies
Describe how a class map is used to define a class of traffic and list which classification
options exist and explain the purpose of packet marking and describe how a policy map is
used to implement traffic marking policy
Identify the Cisco IOS commands that are used to configure and monitor classification and
class-based marking
Explain QoS trust boundaries and explain their significance in LAN-based classification
and marking
Describe data link la\cr-to-nctwork layer interoperability belween different QoS markers
and explain how to configure the mapping
Explain the purpose ofthe link efficiency mechanisms and their functions
Define link categories and explain when link efficiency mechanisms are mandator.
Explain VoIP susceptibility to increased latency when large packets such as FTP transfers
traverse slow WAN links, and specify what serialization delays are generally acceptable for
voice
Describe LFI and explain how to calculate recommended fragment size
Describe how to configure and monitor MLP LFI
Explain how to configure and monitor ERF. 12
Explain how to configure and monitor cR'I'P
*
Modular QoS CLI
fhis topic describes how to implement agiven QoS policy using the MQC.
Modular QoS CU
The MQC provides a modular
approach toconfiguration of
QoS mechanisms.
First, build modules defining
classes of traffic.
Then, build modules defining
QoS policies and assign
classes to policies.
Finally, assignthe policy
modules to interfaces.
Interface Y
The MQC was introduced to allow any supported classification to be used with any QoS
mechanism.
The separation of classification from the QoS mechanism allows new Cisco IOS Software
2"tointroduce new QoS mechanisms and reuse all available classification opttons. On
the other hand, old QoS mechanisms can benefit from new class.ficatton options.
Another important benefit ofthe MQC is the reusability of configuration. MQC allows the
si QoS policv to be applied to multiple interfaces. The MQC, therefore, is aconsolidation of
all the QoS mechanisms that have so far only been available as standalone mechanisms.
Example: Advantages of Using MQC
Configuring committed access rate (CAR), for example, requires entire ^fiB^^
repeated between interfaces and time-consuming configuration modifications. MQC allows the
same QoS policy to be applied to multiple interfaces.
MQC Components
6-64
aa^st-jrasr'^'
MQC Components
Define Classes
of Traffic
"What traffic do we
care about?'
Each class of traffic
is defined using a
class map.
Define QoS Policies
for Classes
"What will be done to this
traffic'?"
Definesa policy map, which
configures Ihe QoS features
associated with a traffic class
previouslyidentifiedusing a
class map.
Apply a Service
Policy
"Where will this
policy be
implemented?"
Attachesa service
policy configured
with a policy map to
an interface.
Follow these steps to implement QoS by using the MQC:
Step t Con figure classification by using the class-map command.
Step 2
Step 3
Configure traffic policy by associating the traffic class with one or more QoS
features using the policy-map command.
Attach the traffic policy to inbound or outbound traffic on interfaces, subinterfaces
or \ irtuul circuits by using the service-policy command.
Class maps are used to create classification templates that are later used in policy maps in
which QoS mechanisms are bound to classes.
Routers can be configured with alarge number of class maps.
You can create aclass map by using the class-map global configuration command. Class maps
are identified bv ease-sens,t,ve names. Fach class map contains one or more conditions that
determine it thepacket belongs to the class.
There are tuo wavs of processing conditions when there is more than one condition in aclass
map.
Match all: This is the default match strategy. All conditions have to he met to bind a
packet lo the class.
Match any: At least one condition has to be met lo bind the packet to the class.
The policy-map command is used to create atraffic policy. The purpose of atraffic policy is to
configure the QoS features thai should be associated with the traffic that has been classified in a
user-specified traffic class or classes. Atraffic policy contains three elements: acase-sensitive
name, atralfic class (specified with the class command), and the QoS policies
The name ofa traffic policy isspecified inthe policy-map CLI (for example, issuing the
policy-map classl command would create atraffic policy named classl). After issuing the
policy-map command, you are placed into policy-map configuration mode. You can then enter
the nameof a trafficclass, and enter the QoS features to applylo the trafficthat matchesthis
class.
The MQC doesnot necessarily require thatyouassociate onlyonetraffic classto a single
traffic policy. When packets match more than onematch criterion, multiple tralfic classes can
be associated with a single traffic policy.
Note Apacket can matchonlyone traffic class within a traffic policy. Ifa packet matches more
than one trafficclass in the traffic policy,the first trafficclass that is defined in the policywill
be used.
The lastconfiguration stepwhen configuring QoS mechanisms using theMQC is usethe
service-policycommand to attacha policymapto the inboundor outboundpackets.
Using thesen ice-policy command, youcanassign a single policy maptomultiple interfaces or
assign multiple policy maps toa single interface (amaximum of one ineach direction, inbound
and outbound). A service policy can be applied for inbound or outbound packets.
Example: Configuring MQC
Consider this example of configuring MQCon a network with voice telephony:
Step1 Classify trafficas voice,high-priority, low-priority, and browserin a class map.
Step 2 Builda singlepolicymapthat definesthree different trafficpolicies(ditTerent
bandwidth and delay requirements for each traffic class): "NoDelay," "BestService,"
and "Whenever," and assign the already defined classes of traffic to the policies.
Voice is assigned to '"NoDelay."High-priority traffic is assigned to "BestService."
Both low-priority and browser traffic is assigned to "Whenever."1
Step 3 Assign the policy map to selected router and switch interfaces.
) 2010 Cisco Systems, Inc. Quality of Service
Configuring Classification
This topic describes the purpose of packet classification.
The component of a QoS feature that recognizes and
distinguishes between different traffic streams
Most fundamental QoS building block
A QoS service class is a logical grouping of packets that
are to receive a similar level of applied quality.
AQoS service class can beany of these:
< Single user: MAC address, IP address...
Department, customer: Subnet, interface...
Application: Port numbers, URL...
Without classification, all packets treated the same
Preparatory stage for marking
Classification is the process of identifying traffic and categorizing that traffic into different
classes. Packet classification uses a traffic descriptor to calegori/e a packet within a specific
group in order to define that packet. Typically used traftic descriptors include CoS. incoming
interface. IP precedence. DSCP. source or destination address, application, and Multiprotocol
label Switching (MPl.S) experimental bits(KXP bits). Alter the packet has been defined (that
is. classified), the packet is then accessible for QoS handling on the network.
Using packet classification, you can partition network traffic into multiple priority levels or
classes of service. When traffic descriptors are used to classify traffic, the source agrees to
adhere to the contracted terms and the network promises a QoS. Different QoS mechanisms,
such as traffic policing, traffic shaping, and queuing techniques use the traffic descriptor ofthe
packet (that is. the classification ofthe packet) to ensure adherence lo thai agreement.
Classification should take place at the network edge, typically in the wiring closet, in IP
phones, or at network endpoints. It is recommended thai classification occurs as close to the
source ofthe traffic as possible.
3-66 Implementing Cisco Voice Communications and QoS (CVOICE) vS.O 2010 Cisco Systems. Inc.
MQC Classification Options
This topic describes the IP packet elassilication options available in MQC.
MQC Classification Opt ons
Classification options configured in a class map
- Requires a referring policy map to be useful
MQC classification options include the following:
- Access list
- IP precedence value
- IP DSCP value
- QoS group number
- MPLS experimental bits
- Protocol (including NBAR)
- Using another class map
- IEEE 802.1Q/ISL
CoS/priority values
- Input interface
Source MAC address
- Destination MAC address
RTP (UDP) port range
- Any packet
- Frame Relay DE bit
Classification using MQC is accomplished by specifying a traffic match criteria within a
configured class map foreach different service class. Inorder for QoS mechanisms touse the
class map. themap must bereferenced through theuse of a policy map, which issubsequently
applied to aninbound or outbound interface as a service policy.
MQC classification withclass maps is extremely flexible and you canclassifypacketsby using
these classification tools:
Access control lists (ACLs): ACLs for any protocol can be used within the class map
configuration mode. The MQC can be used for other protocols, not only IP.
IP precedence: IP packets canbeclassified directly byspecifying IPprecedence values.
DSCP: IPpackets canbe classified directly byspecifying IPDSCP values. Differentiated
serv ices (DilTServ)-enabled networks canhave up to 64 classesif DSCPis usedto mark
packets.
MPLS experimental bits: Packets canbe matchedbasedon the value inthe experimental
bits ofthe MPLS header of labeled packets.
QoS group: A QoSgroupparameter can be used to classify packetsin situations whereup
to 100 classes are needed or the QoS group parameter is used as an intermediate marker.
Forexample, theQoS group parameter canbe used inan MPLS-to-QoS-group translation
on input and a QoS-group-to-DSCP translation on output. QoSgroupmarkings are local to
a single router.
Protocol: Classification is possible by identifying Layer 3 or Layer 4 protocols. Advanced
classification is also available by using the Network-Based Application Recognition
(NBAR) tool, which identifies dynamic protocols by inspecting higher-layer information.
12010 Cisco Systems, Inc. Quality of Service 6-67
Class map hierarchy: Another class map can be used to implement template-based
configurations.
Frame Relay DF. bit: Packets can be matched based on the value ofthe underlying Frame
Relav DL bit.
CoS: Packets can be matched based on the information that is contained inthe threeCoS
bits (when using 802, IQencapsulation) or priority bits (when using Inter-Switch Link
[1SL] encapsulation).
Input interface: Packets can beclassified based onthe interface from which they enter the
Cisco IOS dev ice.
MAC address: Packets can be matched based on their source or destination MAC
addresses.
Lser Datagram Protocol (CDP) port range: Real-Time Transport Protocol (RTP)
packets can be matchedbased on a rangeof UDP port numbers.
AM packets: MQC can also be used to implement a QoS mechanism for all traffic in which
case classification will put all packets into one class.
Field: You can use the match field command to configure the match criteriafor a class
map based on the fields that are defined in protocol headerdescription files (PIlDFs).
Before configuring this match criterion, you must load a PMDF onto the rouler.
Frame Relay data-link connection identifier (DI.CI): You can use the match fr-dlci
command to specify the Frame Relay DI.CI number as a match criterion ina class map.
This match criterion can be used inmain inlerfaces and point-to-multipoinl subinterfaces in
Frame Relav networks. It can also be used in hierarchical policymaps.
MPLS EXP bit value in the topmost label header: You canusethe match mpls
experimental topmost command to matchthe MPLS LXPvalue inthe topmost label. You
can use this match criterion on the input and output interfaces. Il will match onlyon MPLS
packets.
Packet length: You can use the match packet length command to specifythe I.aver3
packet length in the IPheaderas a match criterion ina class map.
Port type: You canuse the match port-typecommand to match traffic based on thepon
type for a class map.
6-68 Implementing Cisco Voice Communicationsand QoS (CVOICE) v8 0 2010 Cisco Systems, Inc.
Class Map Matching Options
fhis subtopic explains the available class map matching options.
Class Map Matching Options
Match all requires all conditions to return a positive answer. If one
condition is not met, theclassmap will return a "no match result.
Match any requires atleast one condition to return a positive _
answer. If no condition ismet, the class map returns a"no match.
The figure illustrates the process of determining if apacket belongs to aclass (match) or not
(no match).
'fhe process goes through the list of conditions and returns the following:
Amatch results ifone ofthe conditions is met and the match-any strategy is used.
Amatch results ifall conditions are met and the match-all strategy isused.
No match results if none ofthe conditions are met.
Quality of Service
Configuring Classification with MQC
6-70
packedMf^ the CiSC I0S CmmandS lhat ^ U"d l Cnn8Ure c,assificat <*
Configuring Classification with
router(configj#
olaaa-map tmatch-any | match-all] claaa-map-name
Enters the class map configuration mode.
Names canbea maximum of 40 alphanumeric characters.
match-all isthe default matching strategy.
router(config-cmap)#
I match condition
Uses at least one condition to match packets.
router[conflg-cmap)#
match class-map clasa-map
One class map canuseanother class map for classification.
Nested class maps allow generic template class maps to be
used in other class maps.
Vou can use the class-map global configuration command to create aclass map and enter the
class map configuration mode. Aclass map is identified by acase-sensitive name. All
subsequent references to the class map must use exactly the same name.
You can use the match command to specify the classification criteria when in class map
configuration mode. You can use multiple match commands within aclass map. At least one
match command should be used within the class map configuration mode. The default is
match none.
You can also nest class maps in MQC1 configurations by using the malch class-map command
within the class map configuration. By nesting class maps, the creation ofgeneric classification
templates and more sophisticated classifications are possible.
Implementing Cisco Voice Communications andQoS(CVOICE) v8 0
Configuring Classification with MQC (Cont)
[match not match-criteria
1
The not keyword inverts the condition.
[match any |
>The any keyword can be used to match all packets.
(match access-group {number | name} [name] |
Attaches an ACL to a class map.
class-mi p Hell-known services
match accasa-group 10 0
class-map Xll-sarvic SB
match any
access-list 10 0 permLt tcp any any lt 1024
These are additional options that give extra power to class maps:
Any condition can be negated by inserting the keyword not.
The any keyword can be used to match all packets.
ACLs are one ofthe most powerful classification tools, Class maps can use any type of ACL
(not only IPACLs). Tlie match access-group command is usedlo attachan ACLto a class
map.
The example in the figure shows these two class maps:
Class map Well-known-services uses an ACL to match all the packets with the source or
destination port number lower than 1024.
Class map All-services actually matches all the packets.
2Q10Cisco Systems, Inc. Quality of Service
Configuring Classification Using Input Interface and RTP Ports
This subtopic deseribes how toclassify traffic that isbased on the input interface and RTP port
range.
Configuring Classification U;
Interface and
router{config-cmap)#
match input-interface interface-name
All packets received through the selected input interface are
matched by this class map.
match ip rtp starting-port-nmnber port-range
Matches UDP packets with source or destination port numbers
within the specified range.
Range is between the starting port (values from 2000 to 65535) and
the sum of the starting port and the port-range (values from 0 to
16383)
class-iap match-any Past Ethernets
match input-interface PaetEtherr etl/0
match input-interface PastEtherr etl/1
e laaH -map RTP
match ip rtp 1S3B4 163B3
fhe malch input-interface command classifies packets based on the input interface.
The match ip rtp command can be used to match RTP packets within a specific UDPport
range.
In the first class-map example, called ""FastHthernets." the malch input-interface will match
am packet that ani\es on either the Fasti-thernell/0 or FaslBhernetl/1 interfaces. In the
second class map. called "RTP." UDP packets in port range starting with port 16384 and
consisting of 16383 ports will be matched.
Configuring Classification Using Marking
This subtopic explains how toclassify packets based onpreviously set marking.
Configuring Classification Using Marking
match cos cos-value [cos-value cos-value cos-value]
match ip precedence ip-prec-vaiue [ip-prec [ip-prec ]]
match [ip] dscp ip-deep-value [ip-deep-value ...]
First command classifies using Layer 2 CoS (802.1 Q or ISL):
- Select up to four CoS or priorityvalues. Range 0 to 7,
* Second command classifies based on the IP precedence
- Matches any packet with one of four IP precedence values
Third command classifies using DSCP
- Matches packet with any of the specified DSCP values
class-map Low-priority
match cob 0 1 2 3
1
class-map VoIP
match ip precedence 5
i
class-map Voice
match ip dscp ef en5
According to the DiffServ model, the per-hopbehavior(PHB)appliedto transit traffic is
identified based on the markers that are set at the network edge. The most common markers are
CoS. IPprecedence, and DSCP. The figure illustrates the syntax ofthe respective match
commands. Theyallowthe matching of any specifiedvalue if multiple numbers are specifiedin
The configuration example includes threeclass maps. Hach matchingis basedon a ditTerent
type of marker.
)2010Cisco Systems, Inc. Quality of Service 6-73
Configuring Class-Based Marking
This topic describes how to mark traffic at the source and remark in transit.
Marking
The QoS feature component that "colors" a packet (frame) so
that it can be identified and distinguished from other packets
(frames) in QoS treatment.
* Packets can be marked with one of these:
IP precedence
IP DSCP
QoS group
- MPLS experimental bits
IEEE 802.1Q, or ISLCoS or priority bits
Frame Relay DE bit
ATM CLP bit
Marking is related to classification. Marking allows network devices to classify a packet or
frame based on a specific traffic descriptor. Typically used traffic descriptors include CoS,
DSCP. IP precedence, and MPLS LXP bits. The Frame Relay discard eligible (DE) bit and
ATMcell loss priority (CLP) have become less common because service providers and
enterprises have been replacing Frame Relay and ATM infrastructure by other transmission
technologies. Markingcan be used to set informationin the I.aver 2 or I aver 3 packet headers.
Markinga packet or frame with its classification allows network devices to easily distinguish
the marked packet or frame as belonging to a specific class. Aller the packets or frames are
identified as belonging to a specific class. QoS mechanisms can be uniformly applied to ensure
compliance with administrative QoS policies.
6-74 Implementing Cisco Voice Communications and QoS (CVOICE) v8 i 2050 Cisco Systems, Inc
Class-Based Marking Overview
fhis topic describes the MQCclass-based marking mechanism.
Ciass-Based Marking Overview
Class-based allows static per-class marking of packets.
- Can mark inbound or outbound packets
- Can be combined with any other QoS feature on output
- Can be combined with class-based policing on input
Cisco Express Forwarding is required on the interface before
the class-based packet marking feature can be used.
Marking packets or frames sets infonnation in the Layer 2 and Layer 3 headers of a packet so
that the packet or frame can be identified and distinguished from other packets or frames.
MQC provides packet-marking capabilities using class-based marking. MQC is the most
flexible Cisco IOS marking tool, extending the marking functionality of committed access rate
(CAR) and policy routing.
Class-based marking can be implemented on input or output interfaces as part of a defined
input or output service policy. On input, class-based marking can be combined with class-based
policing, and on output with the class-based weighted fair queuing (CBWFQ).
Quality of Service
Configuring Class-Based Marking
This topic describes the Cisco IOS commands that are required lo configure class-based
marking.
Configuring Class-Bast
router(config)#
policy-map policy-map-name
Creates a policy map and enters policy map configuration mode
router(config-pmap-cl#
set cos cos-value
set ip precedence ip-precedence-value
set [ip] dscp ip-dscp-value
set mpls experimental mpls-experimental-value
Marks packets in traffic class using CoS. IP precedence, OSCP, or MPLS
EXP
CoS option available for interfaces with ISU802 1Q encapsulation
router (config-if) #
service-policy {input j output} policy-map-name
Associates the policy map to an input or output interface
Marking is configured usingthe set command in policymapconfiguration mode. The four
most common set commands are shown in the figure.
Whenconfiguring class-based marking, these three configurationsteps are required:
Step 1 Create a class map.
Step 2 Create a policy map.
Step 3 Attach Ihe policy map to an interface by using the service-policy command.
ImplementingCisco Voice Communicationsand QoS (CVOICE) v8.0
Class-Based Marking Configuration Example
This subtopic provides a class-based marking configurationexample.
Class-Based Marking Configuration Example
class-nap RTP subnet 10 1 1
rial* Rl nbnt 1C J
I 4lEl> af
iccasa-list 100 permit udp 10.1.1.0 0.0.0.255 range 876S 35000
my ranga 8766 35000
interface FaatEthernet 0/0
fhis figure illustrates a sample configuration for marking RTP traffic with a DSCP value of 46
to ensure the EF PHB. This setting enables timely forwarding of VoIP media in a network with
a DiffServ model, The marking is applied to UDP flows in the defined range, arriving on the
Fast Ethernet 0/0 interface in the incomingdirection.
Quality of Service
Trust Boundaries
This topic describes the concept of a trust boundary and how to implement it in a network.
Trust Boundary Classification
Frame CoS may or may not be trusted by the network.
Classification close to the edge improves scalability.
End hosts cannot be trusted to tag a packet priority correctly.
The outermost trusted device represents the trust boundary.
1 and 2 are optimal, 3 is acceptable.
Endpoints Core WANAgg.
I TrustBoundary
fhe concept of trust is important and integral to deploying QoS. Alter the end devices have set
CoS or ToS values, the switch has the option of trusting them. If the switch trusts the values, it
docs not need to reclassify: if the switch does not trust the values, it must perform
reclassification for the appropriate QoS.
The notion of trusting or nol trustingforms the basis for the trust boundary. Ideally,
classification should he done as close to the source as possible. If the end device is capable of
performing this function, the trust boundary for the network isal the end device. Ifthe device is
not capable of performing this function, or thewiring closet switch doesnot trust the
classification that is done b\ the end device, the trust boundary' might shift.
Howthis shift happens depends on the capabilities ofthe switch inthe wiringcloset. If the
switch can reclassify the packets. Ihetrust boundary is inthe wiringcloset. If the switch cannot
reclassify thepackets, thetask falls to otherdevices in the network, going toward thebackbone.
Inthiscase, onegood ruleis to perform reclassification al thedistribution layer, which means
thatthetrust boundary shirts to thedistribution layer. It is likely that there is a high-end switch
in the distribution la\er with features to support the reclassification function. If possible, try to
avoid performing thereclassification function in thecoreofthe network.
wn
Trust Boundary Marking
This subtopic describes the packet marking performed at the trust boundary with aCisco
Unified IP phone.
Trust Boundary Marking
For scalability, marking should be done asclose to the
source as possible.
Personal Computer
Frames are typically
unmarked (CoS - 0)
unless NIC is
802 1p- of
B02 1Q-capable
- If marked. IP phone
can (and By default
does) reclassify
CoS But riot DSCP
IP Phone
Access Layer
Distribution Layer
Marks voice as Layer 2
CoS (default]or Layer3
ToS or DSCP
Reclassifies incoming
PC data frames
> Typically
- Voice.
- CoS = 5
ToS = 5
. DSCP = EF
- PC
Reclassify
CoS-0
Based on
switch
capabilities
Accept or
remap here
S
Example:
Catalyst 6000
Mark traffic
Accept CoSjToS
Remap CoS to
ToS or DSCP
Classification should take place at the network edge, typically in the wiring closet or within
video endpoints or IP phones themselves.
The figure shows an example of IP telephony packet marking. Packets can be marked using
Layer 2CoS settings, IP precedence, or DSCP. Cisco Unified IP phones can mark voice
packets as high priority using CoS and DSCP. By default, the IP phone sends ^p-tagged
packets with the CoS set to avalue of 5and the DSCP set to Expedited Forwarding (46).
In aCisco IP telephonv environment. PCs are placed in anative VLAN, meaning that their
Ethernet packets are untagged. This means that Ethernet frames originating from aPC will not
have an 802 lp field and thus no provision to set CoS values. By default, DSCP values or
packets originating from PCs are set to 0. Even ifthe PC sends tagged frames with aspecific
CoS value. Cisco Unified IP phones, by default, set the CoS values to zero before sending the
frames to the switch.
Configuring Trust Boundary
This subtopic explains how to configure the trust boundary on aCisco Catalyst switch.
Configuring Trust Boundai
switch]config-If)#
mis qos trust [cos [pasa-through dscp]
mis qos trust device cisco-phone
dscp]
* Configures the port totruststate onan interface
If CoS is trusted'
CoS is used toselect the ingress and egress queues
No pass-through DSCP.
- DSCP modified according to the CoS-to-DSCP mapping
Pass-through DSCP.
Original DSCP retained from ingress to egress
If DSCP is trusted, the DSCP fieldis retained.
CoS modified according totheDSCP-to-CoS map
For non-IP packets CoS setto0. and DSCP-to-CoS map not applied
device cisco-phone enables the Cisco Discovery Protocol trusted boundary
feature *
Otherwise, disables the trusted setting onthe switch port to prevent misuse of
the pnonty queue
The mis qos trust command delines the type oftrust that aCatalyst switch has for trafilc
arm ingon a specific interface. By default, there is no trust.
If Layer 2CoS is trusted (mis qos (rust cos), the CoS marking ofthe incoming packets is used
toselect the ingress andegress queues. Two situations canarise:
If the pass-through dscp option is not configured, the DSCP value in the incoming packet
isoverwritten, using the CoS-to-DSCP mapping table.
The pass-through dscp option causes the original DSCP to be retained in the packet and
betransmitted when thepacket leaves theswitch.
If DSCP is trusted, the DSCP field is retained and not overwritten by the CoS-to-DSCP
mapping table. Instead, the CoS is modified according to the DSCP-to-CoS mapping table. For
non-IP packets. CoS is set to 0and the DSCP-to-CoS map isnot applied.
The device cisco-phone command enables the Cisco Discovery Protocol trusted boundary
feature, which detects ifaCisco Unified IP phone isconnected lo the port. Ifnot, the command
disables the trusted setting on the switch port to prevent misuse ofthe priority queue.
6-80 Implementing Cisco Voice Communications andQoS (CVOICE) v8.
Trust Boundary Configuration Example
This subtopic provides an example of atrust boundary configuration.
Trust Boundary Configuration Example
Traffic sent from the IP
phoneto the switch is
trusted to ensure that
voice traffic is properly
prioritized.
ml8 qos
interface FaatethsrnetO/l
description To Phonal
awitchport mode acceaa
ml8 qos trust cos
ml a qos trust device claco-phoi
avitchport voice vian 110
switchport access vian 10
interface Fastethernet0/16
description To Distribution Svi
mis qos trust dscp
Native VLAN (PVID);
No Co nfigu*cm
Changes Needed on PC
The figure shows atypical connection of aCisco Unified IP phone to aswitch port. Tralfic that
is sent from the telephone to the switch is marked with atag that uses the 802.1Q header. The
header contains the VLAN information and the CoS three-bit field, which determines the
priority ofthe packet. Usually, the switch is configured to trust the marking ofthe voice traffic,
which'is achieved by the mis qos trust device cisco-phone and the mis qos trust cos interface
configuration commands. The upstream port (Fast Ethernet 0/16) is configured to trust the
DSCP marking oftraffic arriving from the distribution layer, which is typical for DiffServ
deployment.
Quality of Service
Mapping CoS to Network Layer QoS
Thistopic deseribes mapping between the data link layer CoS and network layer QoS.
ToS Bitscan be mapped ToS tins can also be
to CoS Brts and vice mapped to MPLSEXP
| versa bitsand viceversa
EXP
IP headers are preserved end-to-end when IP packets are transported across anetwork. Data
link layer headers are not preserved. This means that the IP layer is the most logical place to
mark packets for end-to-end QoS. However, there are edge devices that can only mark frames
at the data link layer, and there are mam other network devices that only operate at the data
link layer. 1oprovide true end-to-end QoS. the ability to map QoS marking between the data
link layer and the network laver is essential.
6-82 Implementing CiscoVoice Communications and OoS (CVOICE) v8 i
Default LAN Switch Configuration
Thissubtopic explains thedefault I.AN switch behavior withregard totrust andmapping of
CoS and DSCP values.
Default LAN Switch Configuration
Typical default values on Cisco Catalyst switches
(may differ on some models):
Port CoS value; 0
Port trust state: untrusted
CoS and DSCP values set for all incoming packets: 0
Different settings for ingress and egress queues
CoS assignment can be modified
By default. QoS is disabled on Cisco switches. When QoS is disabled, there is no concept of
trusted or untrusted ports because the packets are not modified. The CoS, DSCP, and IP
precedence values in the packet are not changed.
When QoS is enabled on a switch, the default port trust slate on all ports is untrusted. As a
result, the switch resets the CoS and DSCP marking for all incoming packets to 0. This
behavior can be changed.
Mapping CoS and IP Precedence to DSCP
This subtopic explains the mapping between data link layer CoS, Layer 3 IP precedence, and
Laver 3 DSCP.
appinq CoS and IP
During classification, configurable mapping tables are used:
To derive a corresponding DSCP or CoS value
From a received CoS, DSCP, or IP precedence value
Default CoS-to-DSCP Map
CoS value 0 12 3
DSCP value 0 I-3 24 32 40 48 56
Default IP Precedence-to-DSCP Map
IP precedence value
0 8 16 24 32 40 48 56
When QoS is enabled and a switch port is trusted, most Cisco Catalyst switches provide these
classification options for IP traffic:
Trust the DSCP \alue in the incoming packet by configuring the port to trust DSCP. for
ports that are on the boundary between two QoS administrative domains, you can modify
the DSCP to another value by using the configurable DSCP-to-DSCP mutation map.
Trust the IP precedence \alue in the incoming packet by configuring the port to trust IP
precedence, and generate a DSCPvalue for the packet by using the configurable IP-
precedenee-to-DSCP map.
Trust the CoS \alue (if present) in tlie incoming packet, and generate a DSCP value for the
packet b\ usingthe CoS-to-DSCP map. If the CoSvalueis not present, use the default port
CoS value.
Perform the classification based on a configured IP standard or an extended ACL. which
examines \arious fields in the IP header. If no ACL is configured, the packet is assigned 0
as the DSCP and CoS value, which means best-effort tralfic. Otherwise, the policy-map
action specifies a DSCP or CoSvalueto assignto the incoming frame.
The figure shows thedefault CoS-to-DSCP and IPprecedence-to-DSCP mapping tables.
CoS-to-DSCP Mapping Example
This subtopic provides an example of CoS-to-DSCP mapping at the ingress network edge.
CoS-to-DSCP Mapping Exampie
DSCP - 40
The figure provides an example of aCoS value that is mapped to the DSCP value in aCisco
Catalyst switch.
The trust boundary has been established on the switch port to trust the CoS setting from the
Cisco Unified IP phone. By default, the phone marks voice traffic with DSCP FF value 46 and
CoS \alue 5.
Because the switch port is configured to trust the CoS setting, the switch uses the CoS-to-DSCP
map to find the appropriate DSCP for the packets. In this case, the map has default settings and
the switch sets the DSCP value to 40. On the switch output, in the Layer 3header, the DSCP
will be set to 40.
Quality of Service
6-85
DSCP-to-CoS Mapping Example
This subtopic pim ides an example of DSCP-to-CoS mapping at the ingress network ed
DSCP-to-CoS Mappm
DSCPvaluas 0 9,10 16, 18 24.26 32 34 4074
CoS values o i j> % a &
CoS = 5 DSCP = 40
Vo'P packet arrives on switch with DSCP 40
Packetprocessedand placedinoutput queue
Switch maps DSCP to CoS
Switch encapsulates packet in 602 1Q. using the mapped CoS value
Switch forwards the frame
The figure shows the prc\ ious packet as it arrives to its destination after traversing the network.
In this example, the ingress port ofthe egress switch is configured to trust DSCP. Therefore
the Layer 3header will have aDSCP value of 40. which was set on the ingress switch. When
the IP packet traverses the egress switch, its outgoing CoS value is set using the DSCP-to-CoS
map. In this example, the map uses default values, and the switch sets CoS 5.
Implementing Cisco Voice Communications and QoS(CVOICE] v80
'2010 Cisco Systems. Inc
turn*
Configuring Mapping
This subtopic explains how to configure mapping between data link layer and network layer
marking.
Configuring Mapping
8witch(config-if)#
mis qos cos {default-cos | override}
Sets CoS for all incoming packets that are untagged
override option overwrites the received CoS. even if port is trusted
switch(config)#
mis qos map cos-dscp dscpl...dscpB
Defines eight DSCP values that correspond to CoS values 0 to 7
Mapping performed only on ports that trust incoming CoS
switch(config)#
mis qos map dscp-cos dscp-list to cos
Maps dscp-list (up to 13 DSCP values) to the corresponding CoS
value (range from 0 to 7)
The mis qos cos interface configuration command defines thedefault CoSvalueon a port. It
assignsa CoSand DSCPvalueto all incoming packetsthat are untagged (if the incoming
packet does not have a CoS value). The override keyword is used lo assign a default CoS and
DSCP value to all incoming packets, even when the frame CoS is set.
The mis qos map global configuration command definesthe CiscoCatalystswitchmapping
between various packet markers. The command includes several options, two of which are
shown in the figure:
CoS-to-DSCP map: This map defines eight DSCPvalues that correspond to CoS values 0
to 7. Mapping is performed only on ports that trust incoming CoS.
DSCP-to-CoS map: Phis setting maps dscp-list (up to 13 DSCP values) to the defined CoS
value (range from 0 to 7).
Other options ofthe mis qos map command, not discussed in this lesson are: the DSCP-to-
DSCP-mutation map. the IPprecedence-to-DSCP map, and the poh'ced-DSCP map.
Quality of Service
Mapping Example
"fhissubtopic provides a sample configurationof CoS-to-DSCPmapping and explains how-
packets are handled by a Cisco Catalyst switch.
Mapping Example
mis qos
mis cos map cos-dacp 0 10 18 26 34 46 48 56
interface Fastethernet0/1
mis qos trust cos
!Map to DSCP using the mapping table
interface FastethernetO/2
switchport mode access
mis qos cob 1
{Untagged frames get CoS=l instead of default 0
mis qos trust cos
[Packets from IP phone are mapped using CoS-DSCP table
TheCatalW switch has QoS enabled, which causes it to reset CoS and DSCPvalues of all
incoming packets, unless configured otherwise, 'fhe mis cosmapcos-dscp command defines
the mapping between incoming CoS values and outgoing DSCP values, for ports that are
configured to trust CoS.
Interface Fast hthernct 0/1 is a trunk port that is configured to trust CoS, so all incoming
802.IQpackets will be subjected totheCoS-to-DSCP map before thepackets aresent out from
the outgoing port.
Interface Fast Lthernet 0/2 is configured to trust CoS, so the taggedpackets sourced from the IP
phone will besubjected tothe CoS-to-DSCP map. Theuntagged packets that aresourced from
the PCattached to the IPphonewill havethe default CoSset to I (withthe nils qos cos 1
command), and therefore their outgoing DSCP will be set to 10.
6-88 ImplementingCisco VoiceCommunications and QoS (CVOICE] v8.0
Link Efficiency Mechanisms Overview
This topic describes the various link efficiency mechanisms and their functions.
Link Efficiency Mechanisms Overview
Link efficiency mechanisms deployed on slowWAN links
(less than 768 kb/s) to:
Increase throughput
- Reduce delay and jitter
* Cisco IOS link efficiency mechanisms include:
- Layer 2 payload compression (Stacker, Predictor, MPPC)
Not effective for VoIP
- Header compression (TCP, RTP, class-based TCP, and
class-based RTP)
- LFI (MLP, FRF.12, and FRF.11.C)
While many quality ofservice (QoS) mechanisms exist for optimizing throughput and reducing
delay in network traffic. QoS mechanisms do not create bandwidlh. QoS mechanisms optimize
the use ofexisting resources and enable the differentiation oftraffic according toapolicy. Link
efficiency QoS mechanisms such aspayload compression, header compression, and LFI are
deployed on WAN links to optimize Ihe use of WAN links.
Compression methods are based oneliminating redundancy. Using header compression
mechanisms, most header information can besent only at thebeginning ofthe session, stored in
a dictionary, and then referenced inlater packets bya short dictionary index. Cisco IOS header
compression methods include TCP header compression, Real-Time Transport Protocol (RTP)
header compression, class-based TCP header compression, and class-based RTP header
compression. Header compression is the most effective method for VolP traffic.
Payload compression isprimarily performed on Layer 2frames and therefore compresses the
entire Layer 3 packet. The Layer 2payload compression methods include Stacker. Predictor,
and Microsoft Point-to-Point Compression (MPPC). Payload compression should not be used
for VoIP.
LFI is a Layer 2technique in which large frames are broken into smaller, equal-sized
fragments, and transmitted over the link in an interleaved fashion with more latency-sensitive
traffic flows (such as VoIP). Using LFI, smaller frames are prioritized and a mixture of
fragments is sent over the link. LFI reduces the queuing delay ofsmall frames because the
frames are sent almost immediately. Link fragmentation, therefore, reduces delay and jitter by
expediting thetransfer of smaller frames. The LFI methods available include MLP andFRF.12.
Link Speeds and QoS Implications
6-90
This topic deseribes themechanisms that should be implemented, depending on the link speed.
Link Speeds and Qc S Smplicati ons
Characteristics
Slow tink
<<768 kb/s)
Medium-Speed
(768-2048 kb/s) (>2048kt)/s) ;
Support for
interactive video
Not
recommended
Yes Yes
LFI Mandatory Not required Not
recommended
cRTP Recommended Optional Not
recommended
Recommended 3-5 classes 3-5 classes 5-11 classes
class model
Fhe table explains which QoS mechanisms should be deployed on WAN interfaces based on
their link speed. Special consideration musl be given tothe slow links (less than 76K kb/s).
Thev should not beused to transport video conferences because they musthavelink
fragmentation and interleav ing. and they should have header compression. Medium speed
inlerfaces may have LFI and cRTP configured, but these mechanisms are not required. Fast
links should not have any link efficiency mechanisms deployed onthem, 'fheclass models
should not exceed five classes onslow and medium speed links and may beextended upto 1
classes onhigh-speed links (more than 2048 kb/s).
Implementing Cisco Voice Communications andQoS (CVOICE) vB.O
Serialization Issues
This topic descnbes VoIP susceptibility to increased latency when large packets traverse slow
WAN links,
Impact of Slow Serial Links on VoIP
Large packets "freeze ouf voice on slow WAN links:
- Excessive delay duetoslow link andlarge packets
Jitter (variable delay) due to vanaWe link utilization
Voice Packet 60
Bytes Every 20 ms
1S0O Bytes
of Data
-214 ms Serialization Delay
10 Mt"s Ethernet
Voice Packet 60 Bytes
Every >214 ms
10 Mb/s Ethernet
When considering delay between two hops in anetwork, queuing delay in arouter must be
considered because it may be comparable to-or even exceed-the serialization and
propagation delay on alink. In an empty network, an interactive or voice session experiences
low or no queuing delav. because the session does not compete wilh other applications on an
interface output queue. Also, the small delay does not vary enough to produce considerable
jitter onthe receiving side.
In acongested network, interactive data and voice applications compete in the router queue
with olher applications. Queuing mechanisms may prioritize voice traffic in the software queue,
but the hardware queue (TxQ) always uses aFIFO scheduling mechanism. After packets of
different applications leave the software queue, the packets will mix with other packes ,n the
hardware queue (TxQ). even iftheir software queue processing was expedited^ heretore, a
voice packet may be immediately sent to the hardware TxQ, where two large FTP packets may
still be waiting for transmission. The voice packet must wait until the FTP packets are
transmitted, thus producing an unacceptable delay in the voice path. Because links are used
variably, the delay varies wilh time and may produce unacceptable jitter in j.tter-sensitive
applications, such as voice.
Serialization Delay
This subtopic explains how to compute the serialization dehr
Serialization Delay
56
64
128
256
512
768
143 us
125us
62 us
31 us
15.5 us
10 us
9 ms
8 ms
4 ms
2 ms
1 ms
18 ms
16 ms
8 ms
4 ms
2 ms
36 ms
32 ms
16ms
8 ms
4 ms
72 ms
64 ms
32 ms
16 ms
8 ms
144 ms 214 ms
128 ms 187 ms
64 ms 93 ms
32 ms 46 ms
16 ms 23 ms
640us 1.28ms 2.56 ms 5.1ms 10.2ms 15ms
Serializationdefay (sec) =
Packet stze in bytes x 8 bits
Link speed in b/s
Example: (1500 bytes x8bits per byte) / 64 kb/s) =187 ms
Serialization delay is the fixed delay that is required to clock avoice or data packet onto the
network interface. Serialization delay is directly related to the link speed and the size ofthe
packet.
fhe figure shows the serialization delay as afunction ofthe link speed and packet size.
For example, the serialization delav for a 1500-byte packet over a56-kb/s link will be ?14 ms
while the serialization delav drops to 15 ms over a768-kb/s link for the same 1500-byte packet.
-92 Implementing Cisco Voice Communications and QoS (CVOICE) v8 0
Link Fragmentation and Interleaving
This topic describes LFI operation and how LFI reduces the delay and jitter of VoIP packets.
Link Fragmentation and interleaving
Splits packets into smaller fragments and interleaves them
with other packets
- Required on slowlinks (< 768 kb/s)
Cisco IOS Software LFI mechanisms include:
- Multilink PPP with interleaving
PPP links
- FRF.12
Frame Relay PVC carrying data traffic, including VoIP
Without LFI
With LFI BE
Fragment Size
I FI isaLayer 2technique, in which all Layer 2frames are broken into small, equal-sized
fragments, and transmitled over alink in an interleaved fashion. LFI reduces delay and jitter by-
expediting transfer ofsmaller frames through the hardware TxQ. LFI should be used on slow
links that do not exceed 768 kb/s.
Thesetwo LFI mechanisms are most commonly implemented in CiscoIOSSoftware:
MLP LFI: By farthe most common andwidely used form of LFI
FRF.12 Frame Relay LFI: Used withFrame Relay dataconnections
The fragments that are transmitted over slow WAN links are self-contained frames, consisting
of I a\ ers 2. 3. and4 headers andpayload. The Layer 2+ headers mustbe considered when
calculating the recommended fragment size.
In particular. Layer 2headers ofFrame Relay with LFI are 8bytes, and Layer 2headers of
MLP with interleaving are 13 bytes.
Fragment Size Recommendation
6-94
fhis subtopic explains how to compute the recommended fragment size based on the link
speed.
Fragment Size Recommendation
Recommended
Link 10 ms I 20 ms 30 ms I 40 ms ["30 ms'T1Dfi'rw|s;2|
mm
56 70 140 210 280 350 700 1400
64 80 160 240 320 400 800 1600
12S 160 320 480 640 800 1600 3200
256 320 640 960 1280 1600 3200 6400
512 640 1280 1920 2560 3200 6400 12800
768 1000 2000 3000 4000 5000 10000 20000
Link speed in b/s
8 bits
Acceptable one-wav delav in a VoIPnetwork is 150ms. Considering the fact that the network
consists of multiple hops andlinks, therecommended serialization delay on a single link should
not exceed 20 ms. This per-Iink delaytranslates, for a givenlink speed, to a maximum frame
size. Because theendpoints transmit data in packets up lo Ihemaximum transmission unit
(MIL1) size, the fragmentation must he perfomied on a per-Iinkbasis.
Implementing Cisco Voice Communications and OoS (CVOICE) v8 0 2010 Cisco Systems, Inc
Configuring MLP with Interleaving
This topicdescribes the Cisco IOScommands that are requiredto configure MLPwith
interleaving.
MLP with Interleaving Configuration Overview
Configuration steps:
Enable MLPon an interface (using a multilink group interface)
Enable MLPinterleaving on the multilink interface
Specify maximum fragment size by setting the maximum delay
on the multilink interface
Follow these steps to configure MLP with interleaving:
Step 1 Lnable MLP on a PPP interface.
Step 2 On the multilink interface, enable interleaving within MLP.
Step 3 In the multilink interface configuration, specify the maximum fragment size by
specifying the maximum desired serialization delay in milliseconds.
i 2010 Cisco Systems. Inc Quality of Service
Configuring MLP with Interleaving
This subtopic describes the Cisco IOS commands that are required to eoniigure MLP wilh
interleaving.
Configuring MLP
router(config-i f)#
ppp multilink
Enables MLP
router(conflg-ifI#
ppp multilink interleave
Enables interleaving of frames with fragments
router(config-i f 1#
ppp multilink fragment delay delay
Configures maximum fragment delay in milliseconds or microseconds
Router calculates the maximum fragment size from the interface
bandwidth and the maximum fragment delay
Fragment size = interface bandwidth * max fragment delay
Default maximum fragment delay is 30 ms
fhe ppp multilink command enables MLP on a PPP interface.
The ppp multilink interleave command enables interleaving of fragments within the multilink
connection.
fhe ppp multilink fragment delay delaycommand specifies the maximumdesired fragment
delay for the interleav ed multilink connection, fhe maximumfragment size is calculated from
the interface bandwidth and the specified maximum delay. The default is set at 30 ms. To
support voice packets, a maximum fragment size of 10 to 20 ms should be used.
fhe ppp multilink fragment delay command specifies the maximumamount of time, in
milliseconds, that should be required to transmit a fragment. If the desired delay should be in
microseconds, set the milliseconds argument lo 0 and enter a value for the microseconds
argument.
6-96 Implementing Cisco Voice Communications and QoS (CVOICE) vS.O 2010 Cisco Systems, Inc.
MLP with Interleaving Example
Thissubtopic provides anexample of MLP with interleaving.
MLP with Interleaving Example
Serial 0/0
Encapsulation PPP
interface Multilinkl
jiyf jl r 111 iik
ppp aultillnk group 1
ppp wilf.lliok EiByMi:'. dais/ 10
ppp Bulbiliok intar*ara
oervice-policy output llq-policy
interface SerialO/0
no ip address
encapsulation ppp
ppp ami
ppp muli
WAN
The figure shows an example configuration of MLPwith interleaving on a multilinkgroup
interface.
A multilink group interface is a collection of interfaces that are bundled together in the
Multilink PPPconfiguration. With a multilink group interface, you can bundle interfaces into
logical multilink groups. The interface multilink command creates a multilink bundle. A serial
interface requires two commands to be assigned to a multilink bundle: the ppp multilink
command, which enables Multilink PPP, and the ppp multilink group command, which
specifies the multilink bundle to which the serial interface should belong.
A maximumdesired delay of 10 ms is configured lo ensure timely forwardingof VoIP traffic.
>2010 Cisco Systems, Inc. Quality of Service 6-97
The show interfaces multilink command output includes MLP LFI statistics infonnation and
indicates whether MI.P interleaving is enabled on the interface. Multilink should be in the open
state along with Link Control Protocol (LCP) and IP Control Protocol (IPCP).
routers show interfaces multilink 1
Multilinkl is up, line protocol is up
Hardware is mul 11". ink group interface
Encapsulation PPP, loopback not. set.
Keepalive set , 1 C sec;
DTR is pulsed for 2 seconds or. reset
L.CP Open, multilink Oper.
Open: IPCP
Input queue : C/",5/ 0/0 : si ze,'max/crops/flushes! ; Total output drops : 0
Queueing strategy: weighted fair
Output q.ieue: C/LOCO/4/0/2441 [size/max total/threshold/drops/interleaves)
Conversations 0/7/16 {active/max active/max total)
Reserved Conversations 0/0 ;allocated/max a 11 ocated)
5 minute input rate 0 bits/sec, 0 packets/sec
5 minute output rate 7C0C bits/sec, 6 packets/sec
The statistics for the output queue include a counter for interleaved frames, which provides a
fair estimate ofthe mechanism effectiveness.
6-98 Implementing CiscoVoice Communications and QoS (CVOICE) v8.0 2010 CiscoSystems. Inc
Configuring FRF.12 Frame Relay Fragmentation
This topic describes when FRF.12 can be used and how FRF, 12 affects VoIP packets.
FRF.12 Configuration Overview
FRF.12 defines fragmentation of Frame Relay data:
Frames that exceed the specified fragmentation size are
fragmented,
- Smaller time-sensitive packets can be interleaved.
- Recommended Frame Relay fragmentation method for VoIP.
FRF.12 requires traffic shaping.
- FRTS or GTS.
- Fragmentation is inactive until traffic shaping is enabled.
FRF.12fragmentation allows longdata frames lo be fragmented into smallerpieces and
interleaved with real-time frames. In this way, real-time voice and nonrcal-time data frames can
be carried together on lower-speed Frame Relay links without causing excessive delay andjitter
to the real-time traffic such as VoIP.
Because Frame Relay is a Layer 2 protocol, it has no way to tell which frame contains voice
(Vol P)or data. Therefore. Frame Relay will fragment all packets larger than the fragment size
intosmaller frames, including VoIPpackets. In a VoIPover FrameRelay network, it is
important to configure thefragment sizeon thedata-link connection identifier (DLCT) so that
VoIP frames willnot get fragmented. Forexample, a G.711 VoIP packetwithout cRTP is 200
byteslong. ForthisDLCI, donot set thefragment sizeto lessthan200bytes.
Cisco IOS Software supports theend-to-end FRF. 12 method, withthe following characteristics:
Packets contain the FRF. 12 fragmentation header.
Fragmentationoccurs at the permanent virtual circuit (PVC) level.
LMI packets are not fragmented.
FRF.12 is configured ona per-PVC basis. Cisco FRF.12 implementation requires thattraffic
shaping is configured (either Frame Relay traffic shaping [FRTS] or generic traffic shaping) for
FRF.12 to be effective. If trafficshapingis not configured, FRF.12is inactive, even when
configured.
Quality of Service
Configuring FRF.12 Fragmentation
fhis subtopic describes the CiscoIOScommands thai are needed to configure FRF.12.
Configuring FRF.12 Fragmentation
router(configl#
map-class frame-relay map-class-name
Specifies a map class to define QoS values fora Frame
Relay VC
routerIconfig-map-class)W
frame-relay fragment fragment-size
* Enables fragmentation for a Frame Relay map class
Sets the maximum fragment size in bytes
routericonflg-if>tt | (conflg-eubl)tt
frame-relay class name
Associates a map class with an interface or subinterface
FRF. 12 fragmentation is configured within the Frame Relay mapclass. The frame-relay
fragment command sets the maximum fragment size in bytes. On an interface, the frame-relay
class command applies the map class to the interface or subinterface. Toassociate a map class
witha DI.CT. use the class command in Frame Relay DLCI configuration modeor Frame Relay
VC-bundle-member configuration mode.
FRF.12 requires FRTSto be enabled.
6-100 Implementing Cisco Voice Communications and OoS(CVOICE) v80
FRF.12 Configuration Example
This subtopic provides an example of FRF.12configuration.
FRF.12 Configuration Example
Serial 0/0
Encapsulation Frame Relay
Serial 0/0.1: DLCI 100
interface serial 0/0
encapsulation frame-relay
interface serial 0/0.1 point-to-point
bandwidth 128
frame-relay interface-dlci 100
class FRF12
1
map-class frame-relay FRF12
IFRTS parameters
frame-relay cir 64000
frame-relay be 2600
frame-relay fair-queue
-Wfthh
Thefigure shows a configuration example where FRF. 12 fragmentation is applied toa data
Frame Relay circuit configured on the serial 0/0.1 subinterface.
Themaximum fragment size is set to 160bytes. Therequired fragment sizeis computed using
this formula:
Fragment_size= Link_speed (b/s) x Delay (sec) / 8 bits
In thecaseofthedesireddelay of 10 ms and link speed of 128,000 b/s, the required fragment
size is 160 bytes. It ensures that VoIP packets, using any audio codec, arenot fragmented.
FRTS is enabled onthe interface andtheFRTS parameters areconfigured within the Frame
Relay map class. Traffic shaping is explained in detail in a later lesson.
The show frame-relay fragment command displaysinformation about the FRF.12Frame
Relay fragmentation process. Thefragment typewill always display end-to-end because this is
the only type that iscurrently supported onCisco IOS Software. Inaddition tofragment type,
the fragment sizein bytes andassociated DLCI is displayed.
router^ show frame-relay fragment
interface dlci frag-type frag-size in-frag
SerialO/0.1 100 end-to-end 80 1298
out-frag
1563
dropped-frag
0
The show frame-relay pvc command output includes settingsthat are relatedto the FRF.12
fragmentation process. This output shows the fragment size (80bytes inthis example) used on
the Frame Relav PVC. The fragment Ivpe is end-to-end.
router> show frame-relay pvc 100
PVC Statistic" for interface SerialO/C (Fiame Relay DTF)
DLC: = ICC, CLCl USAGE - LOCAL, PVC STATUS = INACTIVE, INTERFACE = SerialO/0.1
Cur.puL queue r.ize C/max total 60C./drops 0
fragment type end-r.c-end fragment size 80
cir 6400G be 2600 be 0 limit 325 interval 4C
-mcir j 20cC byte inoreênt 320 BECK response no IF_ CONG no
fraas 153 bytes 15191-'. frags delayed 0 bytes delayed
6-102 implementing Cisco Voice Communications and QoS (CVOICE) v8.0 2010 Cisco Systems, Inc
Class-Based RTP Header Compression
This topic describes class-based RTP header compression.
Header Compression Overview
RTP header compression:
- Maps IP, UDP, and RTP headers to an index
40 bytes (20 + 8 + 12) to 2 or 4 (with checksum)
* Bydefault, 2 bytes (checksum off)
- Used to reduce delay and increase throughput for RTP
- Most effective on slow links (< 768 kb/s)
- Enabled on a link-by-link basis
Two standards-based methods are commonly used to compress headers:
RTP header compression: Usedto compress the packet IP, UDP, and RTP headers, thus
lowering the delay for transporting real-time data, such as voice and video, over slower
links. This method is recommended on slow WAN links carryingVoIPpackets.
TCP headercompression: Also known as the Van Jacobson header compression, TCP
headercompression is usedto compress the packet IPand TCP headersover slowlinks. It
is most effective for small TCP packets, typical for interactive traffic, where the header to
payload ratio is high.
When header compression is enabled, thecompression occurs bydefault inthe fast-switched
pathor theCiscoExpress Forwarding-switched path, depending on which switching method is
enabled on the interface. Class-basedheader compression enables RTPor TCP header
compression on a pcr-elass basis. Decompression is not based on theclassmap. The receiving
end will decompress all packets that come compressed from the other side. Header compression
is performed ona link-by-Hnk basis. Header compression is performed ona hop-by-hop basis
because routers need full Layer 3header information tobeable toroute packets tothe next hop.
RTP Header Compression Example
This subtopic provides an example of R'l'P header compression.
RTP Header Compression Examp!
Link parameters PPP (6 bytes overhead). 64 kb/s
VoIP G 729 (50 samples per second, 20 bytes per sample)
T
45 20
Ovethead = 46 i (46 + 20) = 7D c
Delay = (46 20 .' 64 ktts S = a ms
Band Kith (46 20) 50 8 = 26 4 kb s
2 voice sessions.' 64 lib's
Overhead = 8 / (8 + 20) - 29%
Delay = (8-*-20)/64kb/S-8 = 35ms
Bandwidth = (8 * 20)- 50 * 8 = 11 2 Kb/s
5 voice sessions / 64 kb/s
! PPP r . PPP Layer Z*
<*-: *5m -as.! ~ mn*Tm
G 711 B4 kttfs
G 729 8 kb/s
82 4 ktts
26 4 kb/s
22%
70%
67.2 kb/s
11 2 kb/s
fhe ligure show* ihe packet size before and alter RTP header compression for VoIP packets
using G.729. "fhe IP. UDP. and RTP headers are reduced to2bytes, resulting in 8bytes of
overall headers, fhe VoIP overhead is reduced from 70 to 29 percent. Becauseofthe packet
sizereduction, theserialization delay decreases from 8 ms to3.5ms. 'fhe bandwidth that is
used totransport a single voice call (using theG.729 codec) drops from 26.4 kb/s (66bytes per
frame *50 frames persecond (f/s) *8bits perbyte) to 11.2 kb/s (28 bytes perframe *50f'/s *
8bits per bv te). Therefore, a64-kb/s link can support up totwo G.729 voice calls without
cRTP. but up to fiveG.729voicecalls with cRTP.
When G.7]l-based VoIP is dcploved. theoverhead that is created by the Layer 2 (PPP), IP.
UDP. and RTP headers consumes 18.4kb/s of bandwidth, just as in G.729. fhis overhead
equals 22 percent ofthe overall throughput requirement. 1he overhead can be reduced by using
cRTP. thus consuming 3.2 kb/s of bandwidth, representing 5percent ofthe total eall
bandwidth.
6-104 Implementing Cisco Vorce Communications and QoS(CVOICE) v80
Configuring Class-Based Header Compression
This section describes howto configure class-based header compression.
Configuring Class-Based Header
Compression
router (conflg-paaap-c) #
compression header ip [rtp | tcp ]
Enables RTP orTCP IP header compression:
- In a policy map
- For a specific traffic class
If the rtp or tcp option is not specified, both RTP and TCP
header compressions are configured.
The number of concurrent compressed connections is
automatically determined based on interface bandwidth.
Can be used at any level in the policy map hierarchy
configured with MQC.
Configure class-based TCP and RTP header compression within a policy map using the
compression header ip command. The command can be applied at any level in the policy map
hierarchy that is configured with Modular QoS command-line interface (MQC).
If you do not specify either RTP or TCP header compression, both RTP and TCP header
compressions will be configured.
Note Header compression is autonegotiated only on PPP links. On Frame Relay or High-Level
Data Link Control (HDLC) links, both ends of the links must match.
Class-Based RTP Header Compression Configuration Example
This subtopic provides an example of class-based RTP header compression configuration.
Class-Based RTP Header Compression
Configuration Example
ggSenalO/0 ^
^SHil^ Encapsulation PPP
clasa-map voip
match protocol rtp
policy-map custl
class voip
priority 384
compression header ip rtp
<output omittedi
interface eerialO/0
service-policy output custl
<output omitteda
In the ligure. the compression header ipcommand has been configured to use RTP header
compression for a traffic class called "voip." The voip traffic class is part of a policy map
called"custl." This custt policymap is applied to the sO/0 interface inthe outbound direction.
This poliev provides a maximum bandwidth guarantee of 384 kb/s for the voip traffic class and
will perform RIPheader compression on the voip traffic class leaving the sO/0 interface.
The show policj-map interface command output displavs thetype of header compression
configured (RTP. inthis example), theinterface lo which the policy map called "custl" is
attached (serial 0/0). the number of packets thai aresent, the number of packets that are
compressed, thenumber of bytes saved, andthenumber of bytes sent.
routerfi shov po". icy-map interface Serial 0/0
Service-pel icy cur pur. :custl
Class-map : voip :.matcn-ail i
1005 packers, 64320 bytes
30 second offered rate 16C00 ops, drop rare 0 bps
compress:
header ip rrp
UDP''RTP Compression;
Sent:10D0 roral, 999 compressed,
41957 bytes saved, ;7933 byres sent
3.33 efficiency improvement factor
99% hit raric, five r.m miss rate 0 misses/sec, C max rate 5000 bps
Other statistical infonnation that isprovided inthe output includes the efficiency improvement
factor, which indicates the percentage ofincreased bandwidth efficiency asa result of header
compression. For example, an efficiency improvement factor of 3.33 means 330 percent
effleiencv improvement. The hit ratio isthe percentage ofpackets that are found in the context
database" In most instances, this percentage should be high. The live-minute miss rate isthe
number of trafficflows inthe last five minutes that werenot found inihc contextdatabase. Ihe
rate is the actual traffic rate after the packets are compressed.
6-106 Implementing CiscoVoice Communications and QoS (CVOICEj v80
Summary
Summary
Modular QoS CLI uses class maps to identifytrafficclasses, policy
maps todefine actions, andservice policy toapply thepolicy toa traffic
flow.
Packet classification is a QoS mechanism responsible for distinguishing
between different traffic streams. It can be configured to match various
parameters such as IPheader values, Layer 2 header values, andinput
interface
Packet "colors" a packet so that it can be distinguished fromother
packets traversing the network. Packet marking sets data link layer
values(802.1Q, ISL, MPLS EXP bits, the FrameRelay DE bit, ATM CLP
bit,and IP parameters (IP precedence and DSCP).
Trust boundary demarcates the span of the managed network and is
configured on U\Nswitches using the mis qos trust command.
Mapping ofnetwork layer QoStolink layer CoSandvice versaallows
an end-to-end QoS solution with the marking performed only at the trust
boundary.
Summary (Cont.)
Link efficiency mechanisms include compression (Layer2
payload or header), and LFI.
Low-speed interfaces (below768 kb/s) require link efficiency
methods to provide satisfactory VolP transport.
Link serialization delay should typicallynot exceed 15 ms
(maximumVolP one-way end-to-end delay is 150 ms).
LFI chops large data packets into smaller fragments and
interleaves them with small packets of other flows.
MLP with interleaving requires the configuration of a multilink
interface and does not use MQC.
FRF.12 is configured in combination with trafficshaping.
RTP header compression maps a 40-RTP/UDP/IP header to 2 or
4 bytes. It can be configuredusing MQC ina policy map for a
traffic class.
6-108 [mplementing Cisco Voice Communications and QoS (CVOICE] v8.0 2010 Cisco Systems. Inc
i
Lesson 41
Managing Congestion and
Rate Limiting
Overview
Queuing algorithms are one ofthe primary ways to manage congestion in anetwork Network
devices manage an overflowof arriving traffic by using aqueuing algorithm tolrt traffiZd
determ.ne amethod of prioritizing the traffic onto an output link
Traffic policing controls the maximum rate of traffic that is sent or received on an interface
Traffic^policing ,s used on interfaces at the network edge to limit traffic into or out of the
sTpeefdofthe"S "T^^ "" "^to raalch the Amission rate to the
^ce JXolic.es ^ ^ "* ^ ^^to ^"is^ve quality of
tCWPn?fSed,Weig,hted fr qUeUlng (CBWFQ) Cxte,lds the stiUld^ weighted fair queuing
1 c fUnCtTa'f EVidiB8 SUPPrt fr usei"defi^ traffic classes Aqueu is re rved for
each class, and traffic belonging to aclass is directed lo the queue for that clals
Low latency queuing (LLQ) brings strict priority queuing to CBWFO Strict nrinri,
Tim lesson describes the queuing architecture: traffic-policing, traffic-shaping. CBWFQ, and
Objectives
" "srEeS'i0n' ilS riBin' "* ^ "^ 'br a"*-i êt and rate limiting
6-110
Describe how atoken bucket is used to measure traflic rates and explain token
replenishment and consumption
Fxplain the three models of class-based policing: single token bucket class-based policing.
dual token bucket class-based policing, and dual-rate token bucket class-based policing
Describe how to configure and monitor class-based policing
Describe class-based shaping and explain the two shaping approaches: average rale and
peak rate
Describe how to configure and monitor class-based shaping
Describe LLQ architecture, features, and operation
Identity the MQC commands that arc required to eoniigure and monitor LLQ on aCisco
router
Describe how to calculate voice bandwidth requirements for LLQ conliguralion across
major data link layer technologies
implementing Csco Vo,ce Communications and QoS (CVOICE) v8.0
Congestion and Its Solutions
This topic describes the need for congestion management mechanisms and outlines its
solutions.
Congestion and Queuing: Speed Mismatch
Speed mismatches are the most typical causes of congestion,
Possibly persistent when going from LAN to WAN,
Usually transient when going from LAN to LAN.
. 1000 Mb/s
JfiWf*.
Direction of Data Flow
Congestion can occur anywhere within a network where there are points ofspeed mismatches
(for example, a Gigabit Ethernet link feeding a Fast Ethernet link), aggregation (fbr example,
multiple Gigabit Ethernet links feeding anupstream Gigabit Ethernet link), orconfluence (the
flowing togetherof twoor moretrafficstreams).
Queuing algorithms are used tomanage congestion. Many algorithms have been designed to
serve different needs. Awell-designed queuing algorithm will provide some bandwidth and
delay guarantees to priority traffic.
Speed mismatches arethemosttypical cause of congestion ina network.
Speed mismatches are most common when traffic moves from a high-speed LAN environment
(1000 Mb/s or higher) tolower-speed WAN links (I or2 Mb/s). Speed mismatches are also
common inLAN-to-LAN environments when, for example, a 1000-Mb/s linkfeeds intoa 100-
Mb/s link.
"2010 Cisco Systems, Inc.
Congestion and Queuing: Aggregation
This subtopic explains thesecond common reason forcongestionaggregation.
Congestion and Queuing: Ai
i ip j ___} ___j \_
N * 1 Mb/s
1 Mb/s . 2 Mb/s
Remote
HEHElg]
feg 1000 Mb/s
Chokepoint
\
Chokepoint
Direction of Data Flow
HQ
Thesecond common type of congestion occurs at aggregation points.
In a WAN environment, aggregation congestion istypical on routers that arc attached loWANs
when multiple remote sites Iced back into acentral services site.
In a LAN environment, congestion resulting from aggregation often occurs at the distribution
la>er ofnetworks, where the different access layer devices feed traffic lo the distribution-level
switches.
Queuing Components
This subtopic describes the queuing structure on Cisco router and switch interfaces.
Queuing Components
Congestion:
- Hardware queue is full
- Hardware queue always FIFO
No congestion:
- Packets bypass software queue andgodirectly into hardware queue
Hardware
Queue iTxG!
Queuing on routers is necessary to accommodate bursts when the arrival rate ofpackets is
greater than the departure rate, usually because ofone ofthe following two reasons:
Input interface isfaster than the output interface
Output interface is receiving packets coming in from multiple other interfaces
Thequeuing structure issplit into twoparts, as follows:
Hardware queue: Uses FIFO strategy, which is necessary for the interface drivers to
transmit packets one by one. The hardware queue is sometimes referred to as the transmit
queue (TxQ). Packets in the hardware queue cannot be reordered.
Software queue: Schedules packets into the hardware queue based onthe QoS
requirements. Software queuing is implemented when the interface iscongested. The
software queuing system isbypassed whenever there isroom in the hardware queue. The
software queue is. therefore, used only when data must wait tobe placed into the hardware
queue.
The software queue ismuch larger than the hardware queue, which contains afew packets
(typically 2to 3). The software queue can hold tens ofpackets and allows their reordering prior
to transmission.
The figure illustrates the following actions that have to be taken before apacket can be
transmitted:
Most queuing mechanisms include classification of packets.
After apacket is classified, arouter has to determine whether it can put the packet into the
queue ordrop the packet. Most queuing mechanisms will drop a packet only ifthe
corresponding queue is full (tail drop). Some mechanisms use amore intelligent dropping
scheme, such asweighted random early detection (WRFD).
Ifthe packet is allowed to be queued, it will be put into the FIFO queue fbr that particular
class.
Packet* are then taken from the individual per-class queues and put into the hardware
queue.
Software Interfaces
This subtopic describes congestion management issues on software interfaces, such as
subinterfaces or tunnels.
Software Interfaces
Software-only interfaces cannot perform queuing
- Primarily subinterfaces and tunnels
- Have no concept of departure rate
- No hardware directly tied to them
Solution: hierarchical MQC
* Shaper limits aggregate bandwidth
- Software interfaces mapped to classes within the shaper
MQCPolicy !
Software-only interfaces, such as subinterfaces or tunnels, have no concept of departure rate
because there is nohardware interface that is directly tiedtothem. Nocongestion canoccurand
theycannot perform queuing. Therefore, it is impossible to configure a queuing service policy
directly to a software interface.
Ilierarchical MQC provides a solution byemulating congestion and dividing theforwarding
resources among thesubinterfaces, or traffic classes that arereceived over specific
subinterfaces. Thecongestion is emulated byconfiguring a shaper that throttles thetraffic rate
to a configured value. Class-based queuingis usedto map the subinterfaces or trafficclassesto
individual software queues with bandwidth guarantees. Class-based shaping and queuing are
explained in detail in this lesson.
Policing and Shaping
This topic describes the purpose of traffic conditioning using traffic policing and traffic
shaping.
Policing and Shaping Overview
- These mechanisms must classify packets before policing or shaping the
traffic rate
- Trafficshaping queues excess packets to stay with n the desired traffic rate.
Traffic policingdrops or marks excess trafficto stay withina trafficrate limit.
ail
Transmit, drop, or
mark, then transmit
the packets.
BBB Q
QJ m
BillitB^ÊBF
Buffer
exceeding
packets.
Both traffic shaping and policing mechanisms arctraffic-conditioning mechanisms that are
used in a network to control the traffic rate. Both mechanisms use classification so that they can
differentiate traffic. They bothmeasure the rateof trafficand compare that rate to Ihe
configured traffic-shaping or traffic-policing policy.
The difference between traftic shapingandpolicingcan bedescribed interms of their
implementation.
Traffic shaping buffers excessive traffic sothat the traffic stays within Ihe desired rate. With
traftic shaping, traffic bursts aresmoothed out by queuing the excess traffic loproduce a
steadierflow of data. Reducing trafficburstshelps reducecongestion inthe network.
Traffic shaping is typically used for the following:
To prevent and manage congestion in WAN or MAN networks, where asymmetric
bandwidths arc used along thetraffic path. If shaping is not used, buffering can occur at the
slow (usualK the remote) end. which can lead toqueuing, causing delays, and overflow,
causing drops.
To prevent dropping ofnoncompliant traffic by the service provider. This allows the
customer to keep local control of traffic regulation, but the customer must not exceed the
committed rate.
'I raffle policing drops excess traffic in order to control traffic flow within specified rale limits,
[raffle policing does not introduce any delay to traffic tliat conforms to traffic policies. Traffic
policing can cause more ICP retransmissions, because traffic in excess ofspecified limits is
dropped.
Traffic policing is typically used to satisfy one of these requirements:
Limiting the access rale on an interface when high-speed physical infrastructure is used in
transport. Rate limiting is typically used by service providers to offer customers subrate
access.
Engineering bandwidth so that traffic rates of certain applications or classes of traffic
follow a specified traffic rate policy.
Re-marking excess traffic with a lower priority al Layer 2 and Layer 3, or both, before
sending the excess traffic out. Class-based traffic policing can be configured to mark
packets at both Layer 2 and Layer 3.
Traffic-policing mechanisms such as class-based policing or committed access rate (CAR) also
have marking capabilities in addition to rate-limiting capabilities. Insteadof dropping the
excess traffic, traffic policing can alternatively mark and then send the excess traffic. This
allows the excess traffic to be re-marked with a lower priority before the excess traffic is sent
out. Traffic shapers. on the other hand, do not re-mark traffic; these only delay excess traffic
bursts to conform to a specified rate.
)2010 Cisco Systems, Inc. Quality ofService 6-117
Policing and Shaping Comparison
Thistopicdescribes thedifference between the features of traftic policing andtrafficshaping.
Policing and Shaping Comparis*
Policing
Incoming and outgoing directions
Out-of-profile packets are dropped
Dropping causes TCP retransmits
Supports packet marking or
re-martting
Less buffer usage (shaping requires an
additional shaping queuing system)
Traffic Rate
Policing
Time
Outgoing direction only
Out-of-profile packets are queued until a
buffer gets full
Buffering minimizes TCP retransmits
Marking or re-marking not supported
Shaping supports interaction with Frame
Relay congestion indication
Traffic Rate
Shaping
Time
Shaping queues excess traffic by holding packets inside a shaping queue. Use traffic shaping to
shape the outbound traffic flow when the outbound traflic rate is higher than a configured shape
rate. Traffic shaping smoothes traffic bystoring traffic above theconfigured rate ina shaping
queue. Therefore, shaping increases buffer utilization ona router and causes unpredictable
packet delav s. Traffic shaping can also interact with a frame Relay network, adapting to
indications of Laver 2 congestion in theWAN. Forexample, if thebackward explicit
congestion notification (BLCN) bit isreceived, the router can lower the rate limit tohelp
reducecongestion inthe Frame Relay network.
You can apply policing tocither the inbound oroutbound direction, while you can apply
shaping onlv in the outbound direction. Policing drops nonconforming traffic instead of
queuing the traffic, asshaping does. Policing also supports marking of traffic. Traffic policing
is moreefficient interms of memory utilization than trafficshapingbecause no additional
queuing of packets is needed.
Both traffic policing and traflic shaping ensure that traflic does not exceed a bandwidth limit,
but each mechanism has different impacts on the traffic:
Policing drops packets, generally causing more retransmissions ofconnection-oriented
protocols such asTCP. TCP reacts topacket loss by not only retransmitting the packet, but
also reducing the transmission window size, which defines the number ofoctets that can be
sent unacknowledged. The reduction ofthe window si/e effectively lowers the data
throughput of a TCP session.
Shaping adds variable delav to traflic. possibl> causingjitler.
Measuring Traffic Rates
This topic deseribes how to measure traffic rates.
Single Token Bucket Operations
Outgoing Packet
Be of tokens are added to the
bucket at regular intervals, every Tc.
Number oi tokens
sufficient for
packet stze?
Bucket size = Be + Be (bytes)
Each forwarded packet consumes tokens (equal to packet size).
On average, Be of data is transmitted every Tc.
Occasionally, Be + Be of data can be transmitted if unused tokens have
been accumulated in previous intervals.
Be= Committedburst (amount of daia guaranteed to be deliveredin one time interval)
Be - Exceed burst (additional amount of data attempted to be deliveredif no congestion)
The token bucket isa mathematical model that isused by routers and switches toregulate
traffic flow. The model has the following two basic components:
Tokens: Lach token represents permission to send a fixed number of bils into the network.
Tokensare put into a tokenbucketat a certainrate by CiscoIOS Software.
Token bucket: Atokenbucket has the capacityto holda specifiednumberof tokens. Lach
incoming packet, if forwarded, takes tokens from the bucket, representing the packet size.
If the bucket fills to capacity, newlyarriving tokensare discarded. Discarded tokens are not
available to nature packets. If there are not enoughtokensin the tokenbucket to sendthe
packet, the trafficconditioning mechanisms may take these actions:
Wait for enough tokens to accumulate inthebucket (traffic shaping)
Discard the packet (trafficpolicing)
Using a single token bucket model, the measured traffic rale can conform to or exceed the
specified traffic rate. The measured traffic rate isconforming if there are enough tokens inthe
single token bucket totransmit the traffic, fhemeasured traffic rate isexceeding if there are not
enough tokens in the single token bucket to transmit the traffic.
The committed burst size (Be) isthe amount ofdata that isguaranteed tobedelivered by the
network within onecommitted rate measurement interval (Tc). Il corresponds toa committed
information rate (CIR) using the formula CIR = Be / Tc. Data isalways sent inbursts, and not
evenly with the CIR speed because the frames areputonthewire using theclock rate ofthe
physical circuit. Thus, the committed burst specifies how many octets can besent outin one
interval.
>20lOCisco Systems, Inc.
fhe excess burst (Be) is theamount of datathat thenetwork agrees todeliver lo thedestination
if there is no congestion inthe network. TheBeand Be values arepart ofthe contract that an
enterprise or user signs witha serviceprovider.
Example: Token Bucket as a Coin Bank
You can think of a token bucket as a coin bank. You can insert a coin into the bank every dav
(thetoken bucket). At anv given time, youcanonly spend what youhave saved in thebank. On
the average, if\our saving rate isadollar per day, your long-term average spending rate will be
one dollar per dav ifvou constantly spend what you saved. However, ifyou do not spend any
monev on a iv en dav. vou canbuild upyoursavings inthebank tothe maximum that thebank
can hold. Lor example, if the size of the bank is limited tofive dollars, and if you save and do
not spend for fwc straight days, the bank will contain five dollars. When the bank fills toits
capacitv. \ou v\ill not be able to put any more monev into it. Then, at any lime, you can spend
up to five dollars (bursting above the long-term average rate ofone dollar per day).
The conforming rate (orthe Be) (using the coin bank example) means that ifyou have two
dollars inIhe bank and vou try tospend onedollar, that isconsidered conforming because vou
are not spending more than\ou have saved.
Exceeding rate (or the Be) (using the coin bank example) means that ifyou have two dollars in
the bank and vou try to spend three dollars, il isconsidered exceeding because you are spending
more than \ ou hav e sav ed.
6-120 implementing Cisco Voice Communications and QoS (CVOICE) vS.O 2010 Cisco Systems, Inc.
Single Token Bucket
This subtopic describes the single token bucket operation.
Single Token Bucket
If sufficient tokens are available (conform action):
- Tokens equivalent to the packet size are removed from
the bucket,
-- The packet is transmitted.
500 Bytes
Transmit
TTie figure shows a single token bucket traffic-policing implementation. Starting with a current
capacity of 700 bytes worth of tokens that are accumulated in the token bucket, when a 500-
b\ te packet arrives at the interface, its size is compared to the token bucket capacity (in bytes).
The 500-byte packet conforms to the rate limit (500 bytes < 700 bytes), and the packet is
fonvarded. 500 bytes worth of tokens are taken out ofthe token bucket, leaving 200 bytes
worth of tokens for the next packet.
Single Token Bucket (Cont.)
* Ifsufficient tokens are not available (exceed action):
Drop (or mark) the packet
Token Bucket
200 ;
Bytes ''},
i Remain: J
lig
_D
S L-rop
When the next 300-bvte packet arrives immediately after the first packet, and no new tokens
have been added to the bucket (tokens are added periodically), the packet exceeds the rate limit.
The current packet si/e (300 bvtcs) is greater than the current capacity ofthe token bucket (200
bvtes). and the exceed action is perfomied. The exceed action can be to drop or mark the packet
during traffic policing.
Class-Based Policing
This topic describes class-based policing.
Single Token Bucket Class-Based Policing
Be is normal burst size
Tc is the time interval
CIR is the committed information rate
CIR = Be / Tc
Token Arrival
Rale (CIR)
Token bucket
Excess tokens
are discarded
Token bucket operations rely on parameters such as CIR, Be, and Tc. Be is known as the
normal burst rate. The mathematical relationship between CIR, Be, and Tc is as follows:
CIR (b/s) = Be (bits) / Tc (sec)
With traffic policing, new tokens are added into the token bucket based on the interpacket
arrival rate and the CIR. Every time that a packet is policed, new tokens are added back into the
token bucket. The number of tokens added back into the token bucket is calculated as follows:
(Current Packet Arrival Time - Previous Packet Arrival Time) * CIR
An amount (Be) of tokens is forwardedwithout constraint in every time interval (Tc). For
example, if 8000bits (Be) worthof tokens are placedinthe bucketevery250 ms (Tc). the
router cansteadily transmit 8000bits every250ms if trafficconstantly arrivesal the router.
CIR (normal burst rate) = 8000 bits (Be) / 0.25 seconds (Tc) = 32 kb/s
Without anv excess bursting capability, if the tokenbucketfills to capacity (Be of tokens), the
token bucket will overflow andnewly arriving tokens will bediscarded. Using theexample, in
whichthe CIRis 32 kb/s (Be= 8000bits and Tc = 0.25 seconds), Ihemaximum trafficrate can
never exceed a hard rate limit of 32 kb/s.
Dual Token Bucket Single Rate Class-Based Policing
I his subtopic describes dual token bucket single rate class-based policing.
Dua! Token Bucket Single-Rate, Class
Policinq
Be is the normal burst size
Tc is the time interval
CIR is the committed information rate
CIR = Bc/Tc
Overflow
E^^^tf KaSwIiil
Excess tokens
are discarded.
You canconfigure class-based traffic policing tosupport excess bursting capability. With
excess bursting, after the first token bucket is filled to Be. extra (excess) tokens can be
accumulated in a second token bucket, fxeess burst (Be) is the maximum amount of excess
trafficover and above Bethat can be sent duringthe time interval after a periodof inactivity.
With asingle rate-metering mechanism, the second token bucket wilh a maximum size ofBe
fills at the samerate(CIR) as the first tokenbucket. Ifthe second tokenbucket fills upto
capacitv. no more tokens can beaccumulated and the excess tokens arediscarded.
When using a dual token bucket model, themeasured traffic ratecan beas follows:
Conforming: Ihere areenough tokens inthe first token bucket with a maximum size
of Be.
Exceeding: fhere are not enough tokens in the first token bucket, but there are enough
tokens in the second token bucket vv ith a maximum size of Be.
Violating: There arenot enough tokens inthe first or second token bucket.
With dual token bucket traffic policing, the tvpical actions that areperfomied aresending all
conforming traffic, re-marking (to a lower priority), sending all exceeding traffic, and dropping
all violating traffic. The main benefit ofusing adual token bucket method isthe ability to
distinguish between traffic that exceeds the Be but not the Be. This enables adifferent policy to
be applied to packets in the Be category. Referring to Ihe coin bank example, think ofthe CIR
as the savings rate ($I per day). Be ishow much you can save into the bank per da> ($1). Tc is
the interval at which vou put monev into the coin bank (one day). Be ($5) allows you toburst
over the average spending rate ofone dollar per day ifyou are not spending adollar perdav.
Using adual token bucket model allows traffic exceeding the normal burst rate (CIR) to be
metered as exceeding, and traffic thai exceeds the excess burst rate to be metered as violating^
traffic. Different actions can then beapplied to the confomiing, exceeding, and violating traffic.
6-124 Implementing Cisco Voice Communications and QoS(CVOICE) v80
Dual Rate Class-Based Policing
Dual-Rate Class-Based Policing
Tc: Tokens in CIR bucket
Tp: Tokensin PIRbucket
Enforce traffic policing according totwo separate rates:
- Committed information rate
- Peak information rate
Token Amval
Rate (PIR)
Packet
of StzeB
____ No sJmmM^ No
~^n-^^m- 1
Token Amval
Rate (CIR)
"I
m
With dual-rate metering, the traffic rate can be enforced according to two separate rates. CIR
and peak infonnation rate (PIR). Before this feature was available, you could meter traffic
using asingle rate that was based on the CIR with single or dual buckets. Dual-rate metering
supports ahigher level of bandwidth management and supports asustained excess rate that is
based on the PIR.
With dual-rate metering, the PIR token bucket fills at arate that is based on the packet arrival
rate. The configured PIR and the CIR token bucket fills at arate that is based on the packet
arrival rate and the configured CIR.
When apacket arrives, ihe PIR token bucket is first checked to see if there are enough tokens in
the PIR token bucket to send the packet. The violating condition occurs ifthere are not enough
tokens in tlie PIR token bucket to transmit the packet. Ifthere are enough tokens in the PIR
token bucket to send the packet, then the CIR token bucket is checked. The exceeding condition
occurs ifthere are enough tokens in the PIR token bucket to transmit the packet but not enough
tokens in the CIR token bucket to transmit the packet. The conforming condition occurs if there
are enough tokens in the CIR bucket totransmit the packet.
Dual-rate metering is often configured on interlaces at the edge ofanetwork to police the rate
of traffic entering or leaving the network. In the most common configurations, traffic that
conforms is sent and traffic that exceeds is sent with adecreased priority oris dropped. You
can change these configuration options to suit your network needs.
This token bucket algorithm provides users with three different actions for each packet:
A conform action
An exceed aclion
Anoptional violateaction
Qualityof Service 6-125
3-126
Traffic entering the interface with two-rate policing configured is placed into one of these
categories, within these three categories, users can decide packet treatments. For example a
usermay configure a policing policy as follows: '
Conforming packets are transmitted. Packets that exceed may be transmitted with a
decreased priority: packets that violate are dropped.
The violating condition occurs ifthere are not enough tokens in the PIR bucket to transmit
ttie packet.
The exceeding condition occurs if there are enough tokens in the PIR bucket to transmit the
packet but not enough tokens in the CIR bucket to transmit the packet. In this case the
packet can be transmitted and the PIR bucket is updated to Tp-B remaining tokens where
lpis the si/e of the PIR bucket and Bis the si/e ofthe packet to be transmitted.
The confomiing condition occurs if there are enough tokens in the CIR bucket to transmit
the packet. In this case, the packets are transmitted and both buckets (Tc and Tp) are
decremented to Tp-B and to Te-B. respectively, where Tc is the si/e ofthe CIR bucket Tp
is the si/e of the PIR bucket, and Bis the size ofthe packet to be transmitted
Implementing Cisco Voice Communications andQoS (CVOICEI v8 0
Configuring Class-Based Policing
This topic describes how lo configure class-based policing.
Class-Based Policing Implementation
^^H Feature 1 Dfiscrirrtion ^^^fl
Configuratton
method
Implementations
Conditions
Actions
Multiactjons
MQC
Single or dual token bucket, single or dual rate
Conform, exceed, violate
Drop, set (remark), transmit
Applying two or more set parameters as a conform
or exceed or violate action
The class-based policing feature performs these functions:
Limits the input or output transmission rate of a traffic class that is based on user-defined
criteria
Marks packets by setting different Layer 2 or Layer 3 markers, or both
Class-based policingcan be implemented by usinga singleor doubletoken bucket method as
themetering mechanism. When theviolate action option is not specified in the police MQC
command, the singletokenbucket algorithm is engaged. Whenthe violateactionoption is
specified in the police MQC command, thedual token bucket algorithm is engaged.
Adual token bucket algorithm allows traffic to dothe following:
Conform to theratelimitwhen thetraffic is within the average bit rate
Exceedthe rate limit when the traffic exceeds the average bit rate, but does not exceed the
allowed excess burst
Violate the rate limit whenthe trafficexceedsboth the averagerate and ihe excess bursts
Depending on whetherthecurrent packet conforms with, exceeds, or violatesthe rate limit,one
or more of these actions can be taken: transmit, drop, or set a marker and transmit.
Multiaction policing is a mechanism that canapply more than one action to a packet; for
example, setting thedifferentiated services codepoint(DSCP) as wellas thecell losspriority
(CLP) bit on the exceeding packets.
>2010CiscoSystems, Inc.
This subtopic describes the Cisco IOS Software command that is required to configure single-
or dual-rate class-based policing.
police {cir cir} [be conform-burst] {pir pir} [be peaJr-
burst] [conform-action action] [exceed-action action]
[violate-action action]
Specifies both the CIR and the PIR for two-rate trafficpolicing
CIR - committed information rate (b/s)
PIR = peak information rate (b/s)
The Be and Be keywords and their associated arguments
{conform-burstand peak-burst, respectively) are optional.
be default: 1500 bytes or CIR / 32, whichever is higher
be default: Equal to Be
fhe MQC-based police command defines policing parameters for eithera single- or dual-rate
policer. The cirparameter defines the policed CIR; Be and Be define the token bucket sizes in
bvtes: and the action defines an action for confomiing, exceeding, or, optionally, violating
traffic.
If Be(inbvtcsI is not specified, it will default to CIR/ 32. or 1500 bytes, whichever is higher.
When using the formula CIR / 32 tocalculate the default Re (inbytes), Cisco IOS Software
uses a Tc of 0.25 second, where:
Be (in bytes'' = (CIR
Be (in bvtes' = (CIR
Tc) / 8
0.25 seconds
) /
= CIR / 32
If Be (inbvtes) is not specified, it will default to Be. In a single token bucket case. Cisco IOS
Software ignores the Be value, fhis means that excess bursting isdisabled.
The Beratecanbespecified when a violate action is configured, therefore using a dual token
bucket. This allows Reto be explicitly configured instead of usingthe default valueof Be = Be.
Bespecifies the si/e ofthe second (excess)tokenbucket.
Definition ofthe pir parameter enables dual-rate policing, which uses twoseparate rales: CIR
and PIR. fhe Be and Bekeywords and their associated arguments (conform-burst and peak-
burst, respectivelv) areoptional. If Beis not specified. Be(inbytes) will default toCIR / 32, or
1500 bytes, vvhichever ishigher. If Be isnot specified. Be (in bytes) will default toPIR/32. or
1500 bytes, whichever is higher.
6-128 Implementing CiscoVoice Communications and QoS (CVOICE] vS.O
Class-Based Policing Example: Single Rate, Single Token
Bucket
This subtopic provides aconfiguration example of single rate, single token bucket class-based
policing.
Class-Based Policing Example:
Single Rate, SingleToken Bucket
www.123.com
www.456.com
Police incoming traffic
from the web servers.
-lass-map wwv.123.com
match sourc.-address .= OQOd.dddf.0480
1
rlass-map www.456.ccan
natch sourca-address mac rjrj0d.dddc.od21
policy-map ServerFa
clans www.123.com
class www.456.cob
polica
interface PastBthernet 0/0
Berviaa-policy input SarvecFa
The class-based policing configuration example shows two configured traffic classes hat are
based on upstream MAC addresses. Traffic from the particular web server, which is classified
by its MAC address, is policed to afixed bandwidth with no excess burst capability using a
single token bucket. Conforming traffic is sent as-is and exceeding traffic is dropped In this
case, the ûw.123.com web server is policed to arate of 512 kb/s, and the www.456.com web
server is policed toa rate of256 kb/s.
Because the violate action is not specified, this will use asingle token bucket scheme and no
excess burstingis allowed.
In this example, the Be is not specified, and therefore it will default to 512,000 /32 (16,000
bytes) and 256.000 / 32 (8000 bytes), respectively.
The default Be setting can be examined by showing the policy map. The Be is not displayed
because no excess bursting is allowed using asingle token bucket with class-based policing.
Class-Based Policing Example: Single Rate, Dual Token Bucket
This subtopic prov ides aconfiguration example of single rale, dual token bucket class-based
policing.
Class-Based Policing Example;
Single Rate, Dual Token Bucket
www 123 com
d.dddc.ad21
policy-map ServerParm
claea www.123.com
police 912000 conform- act ion iwk-pkc-**it * *r-ewi-action
<.t-piM-WMt.it 3 Tioios,. action drop
claas wvw.456.com
polic- J5S000 confon.-actioo Mt-prac-transmit 4 axoead-action
set-prac-transmit 3 violate-action drop
interface Fast Ethernet O/rj
service-policy input ServerFarm
The class-based policing configuration example shows two configured traffic classes that are
based on upstream MAC addresses.
Traffic from the particular web server, which is classified bv its MAC address is policed to a
fixed bandwidth with excess burst capability using adual token bucket, by configuring aviolate
action. Conforming traffic will be sent as-is. exceeding traffic will be marked to IP precedence
3and transmitted, and all violating traffic will be dropped.
In this example, because the violate action is specified, adual token bucket scheme with excess
bursting will be used, fhe committed burst si/e (Be) is not specified, and therefore it will
default to the 512.000 / 32 (16.000 bvtcs) and 256.000 / 32 (8000 bytes), respectivelv. The
excess burst si/e (Be) isalso not specified, and therefore itwill default toBe.
Implementing Cisco Voice Communications and QoS(CVOICE) vS.O
Class-Based Shaping
This topic describes class-based shaping.
Two Traffic Shaping Methods
Shaping to Average Rate Shaping to Peak Rate
Characteristics Forwarding packets at the
configured average rate
(Be of trafficat every Tc).
Allowed bursts up to Be
when extra tokens
available.
Recommended Conservative.
use More common method.
Forwarding packets at the peak
rate of up to Be + Be of trafficat
every Tc However, trafficsent
above the CIR may be dropped
during network congestion.
Network has additional
bandwidth available.
Application tolerates occasional
packet loss.
Class-based traffic shaping applies for outbound traffic only. Class-based shaping can be
configured in the following two ways:
Shaping to the configured average rate: Shaping to the average rate forwards up to a Be
of traffic at every Tc interval, with additional bursting capability when enough tokens are
accumulated in the bucket. Be of tokens are added to the token bucket at every Tc time
interval. After tlie token bucket is emptied, additional bursting cannot occur until tokens are
allowed to accumulate, which can occur only during periods of silence or when the transmit
rate is lower than the average rate. After a period of low traffic activity, up to Be + excess
burst (Be) of traffic can be sent. This is the most common method of configuring class-
based shaping.
Shaping to the peak rate: Shaping to the peak rate forwards up to Be + Be of traffic at
every Tc time interval. Be + Be of tokens are added to the token bucket at every Tc time
interval. Shaping to the peak rate sends traffic at the peak rate, which is defined as the
average rate multiplied by (1 + Be / Be). Sending packets at the peak rate may result in
dropping in the WAN cloud during network congestion. Shaping to the peak rate is
recommended only when the network has additional available bandwidth beyond the CIR
and applications can tolerate occasional packet drops.
) 2010 Cisco Systems, Inc. Quality of Service 6-131
Configuring Class-Based Shaping
This topic describes the Cisco IOS commands that are required to configure class-based
shaping.
snfiqurinq Class-Based Shi
router(conflg-pmap-c)#
shape {average
shape {average
peak} average-bit-rate [Be] [Be]
peak} percent [Be] [Be]
Configures shaper in b/s or percent
Recommended to omit the Be and Be to let Cisco tOS
Software select optimal values
The shape average and shape peak commands configure average and peak shaping,
respectivelv. Ihe Be and Be value in bits can be explicitly configured, or Cisco IOS Software
can automatical!) calculate their optimal value. It is not recommended that you configure the
Be and Be in order to let the Cisco IOS algorithm determine the best Be and Be value to use.
Class-basedtraflic shaping uses a single token bucket with a maximumtoken bucket size of Be
+ Be.
The shapepercent command is often used in conjunction withthe bandwidth and priority
commands. The bandwidth and priority commands can be used to calculate the total amount
of bandwidth available on an entity (for example, a physical interface). When the bandwidth
and priority commands calculate the total amount of bandwidth available on anentity, the total
bandwidth is the bandwidth on the physical interface.
Class-Based Shaping Example
"fhis subtopic provides anexample of class-based shaping.
Class-Based Shaping Example
Shaping to 16 kb/s
class-Dap Shape
match protocol citrix
policy-map ShapeAvg
class Shape
ban* #**** 18M0
i
policy-map SbapePea*
class Shape
bap* po*k KOOfl
i
interface SarialO/0
service-policy output ShapeAvg
1
interface SarialO/1
service-policy output ShapePsak
Cisco IOS Software calculated values:
Be = Be = 8000 bits, Tc = 500 ms
Peak rate =Avg. rate * (1 + Be / Be)
= 16000* (1 +8000/8000)
= 32000 b/s
The figure showsan exampleconfiguration for standalone class-based shaping(noCBWFQ).
Citrix traffic is classified into the Shape class.
The Shape class is then shaped to different rates on two interfaces:
On the serial 0/0 interface, traffic is shaped to the average rate. The Be and Be values are
not configured, allowingCiscoIOSSoftwareto automatically calculatetheir optimal
values. The Citrix traffic is shaped to the average rate of 16,000b/s. The resulting
automatically determined Be and Be values will both be 8000 bits with a Tc of 500 ms.
On the Serial 0/1 interface, the Citrix traffic is shaped to the peak rate. Because the Be and
Be are not specified. Cisco IOS Software automatically calculates the optimal value for Tc.
Be, and Be. Ihe shape statement is shape peak 16000. The resulting automatically
determined Be and Be values will each be 8000 bits with a Tc of 500 ms. Therefore, the
peak rate will be as follows:
peak rate = average rate
16000 * 2 = 32000 b/s
* (1 + Be / Be] = 16000 * (1 + 8000 / 8000) =
Hierarchical Class-Based Shaping with CBWFQ Example
This subtopic provides an example of hierarchical class-based shaping with CBWI'Q.
Hierarchical Class-Based Shaping
CBWFQ Example
Moves the bottleneck to the local interface
Tomanage congestion and prevent uncontrolled drop in WAN
Divides the aggregate bandwidth among classes
Can be used to map packets from software interfaces
Child
Policy Map
Parent
Policy Map
clin lubclasa-x'
bandwidth pa
bandwidth pan
clan iubclaia-1
bandwidth pate
50
20
10
|pttficy-iMB .hap*-11J
cilia claaa*-da*faiilt
ahapa avaraga 3B4000
Dtalfm Sorlal 0/0
aivlcai.iftul: ollal
The example uses hierarchical policy maps and configures CBWI'Q inside the class-based
shaping, 'fhe parent policy is the shape-all policy. This parent policy references a child policy
named child-cbwfq. The parent policy map shape-all specifies an average shape rate of 384
kb/s for all the traffic (matched by class-default) and assigns the service policy that is called
child-cbwfq as the child policy.
fhe shaped traffic is further classified into three distinct traffic subclasses, with bandwidth
guarantees of 50. 20. and 10 percent ofthe shaped bandwidth, respectively.
Ihe traffic that is transmitted over the serial 0/0 interface is rate-limited to a total of 384 kb/s.
Suhclass-x has a minimal guarantee of 192 kb/s (384 * 0.5), subclass-y has a guarantee of 76.8
kh/s (384 * 0.2). and subclass-z has a guarantee of 38.4 kb/s (384 * 0.1).
Hierarchical class-based shaping with CBWKQ is used to throttle the transmission rate and thus
manage congestion locally instead of relying on the network to deliver or drop the packets to
the destination. The class-based bandw idth guarantees within the parent shaper ensure a certain
forwarding rate for each class.
Hierarchical class-based shaping with CBWFQ can be used to manage congestion on solluare-
based interfaces, such as subinterfaces or tunnels, because they do not have a hardware queue
that is directlv associated w ith them.
6-134 Implementing Cisco Voice Communications and OoS (CVOICE) vS.O 2010 Cisco Systems. Inc.
Low-Latency Queuing
This topic describes LLQ and CBWFQ.
Class-Based Weighted Fair
and Low-Latency Queuing
CBWFQ:
Mechanism that is used to guarantee bandwidth to classes
- Each class has a reserved queue
- Each class can perform WRED to avoid congestion
- Each class gets more than reserved bandwidth when there is
no congestion
Unused bandwidth allocated proportionally to guarantees
LLQ:
Adds priority queue to CBWFQ for real-time traffic
High-priority class:
- Low-latency propagation of packets
- Guaranteed bandwidth
- Policingto guaranteed bandwidth when congestion occurs
CBWFQ defines multiple traffic classes tliatarebased on match criteria. Packets satisfying the
matchcriteriafor a class constitute the traffic for that class. Aqueueis reservedfor each class,
andtraffic belonging to a classis directed tothat classqueue. Aclassis characterized by
bandwidth, weight andmaximum packet limit. Thebandwidth that is assigned to a class is the
minimum bandwidth that is delivered to theclassduring congestion.
CBWFQ supports multiple class maps toclassify traffic into itscorresponding FIFO queues.
Taildropis thedefault dropping scheme. WRED canbe used incombination withCBWFQ to
prevent congestion of a class.
For CBWFQ. the weight for a packet belonging to a specific class is derived fromthe
bandwidth that you assigned to the class whenyou configured it. Therefore, the bandwidth that
isassigned to tlie packets ofa class determines the order inwhich packets are sent. Aclass gets
more than its reserved bandwidth if there is no congestion. The unused bandwidth is shared
among the classes proportionally to their guarantees.
InCBWFQ. all packets areserviced fairly based onweight; noclass of packets may begranted
strict priority. This scheme poses problems for voice traffic, which is largely intolerant of
delay, especially variation indelay. Forvoice traffic, variations indelay introduce irregularities
of transmissionthat are heard as jitter in the conversation.
The LLQ feature brings strict priority queuing toCBWFQ. Strict priority queuing allows delay-
sensitive datasuch as voice to bedequeued andsent first (before packets inotherqueues are
dequeued), giving delay-sensitive data preferential treatment over other traffic.
LLQ Architecture
This subtopic describes the LLQ architecture.
LLQ Architecture
* Priority queue served first within guaranteed bandwidth
CBWFQ classes protected from starvation by policing of the
priority queue
MQC Classifies!iOl
When CBWFQ is configured as thequeuing system, it creates a number of queues, intowhich
il classifies traftic classes. Thesequeuesare thenscheduled witha weighted scheduler, which
can guarantee bandwidth to each class.
If LLQ is usedwithin the CBWFQ system, it createsan additional priority queuein the WFQ
system, which is serviced by a strict priority scheduler. Any class of IraiTiccan therefore be
attached to a service poliev. which uses priority scheduling, andhence canbe prioritized over
other classes.
Classes towhich the priority command is applied arcconsidered priority classes. Within a
policy map. >ou can assign prioritv status toone or more classes. When multiple classes within
a single policy map areconfigured aspriority classes, all traffic from these classes isqueued to
the same, single, strict prioritv queue.
6-136 Implementing CiscoVoice Communications and QoS (CVOICE) u8.0
LLQ Benefits
This topic describes the benefits of LLQ.
LLQ Benefits
High-priority classes are guaranteed:
tow-latency propagation of packets
- Bandwidth
Consistent configuration and operation across all media types
Entrance criteria fo a class can be defined by an ACL:
- Defines trust boundary to ensure simple classification and
entry to a queue
Supports DiffServ EF and AF PHBs:
- Priority queue implements EF PHB
- CBWFQ with WRED implements AF PHBs
ITie LLQpriority scheduler implements Expedited Forwarding Per-Hop Behavior (FF PHB) of
the DiffServ model by guaranteeing both low-latency propagation of packets and bandwidth to
high-priority classes. Low latency is achieved by expediting traffic using a priority scheduler.
Bandwidth is also guaranteed by the natureof priority scheduling, but is policedto a user-
configurable value. The strict priority scheme allows delay-sensitive data such as voice to be
dequeued and sent firstthat is, before packets in other queues are dequeued.
Policing of priority queues also prevents the priority scheduler frommonopolizing the CBWFQ
schedulerand starvingnonpriority classes, as legacy PQdoes. Byconfiguring the maximum
amount of bandwidth that is allocated for packetsbelonging to a class, you canavoid starving
nonpriority traffic.
The nonpriority class queues within the LLQ (or CBWFQ) structure implement the Assured
Forwarding (AF) PHBofthe DiffServ model by guaranteeing a definedservice level to each
class. The differential drop is controlled by WRED, whichcan be configured individually for
each class, and specifies the drop characteristics ofthe traffic in the class.
Quality ot Service 6-137
Configuring LLQ
This topic describes how to configure LLQ.
Configuring LLQ
priority bandwidth [burst]
priority percent percentage [burst]
* Declares a class within policy map as LLQ class
Allocates a fixed amount of bandwidth (in kb/s) or percentage of configured or default
interlace bandwidth
Trafficexceeding the specified bandwidth is dropped if congestion exists
Default burst size based on 200-ms interval and LLQ bandwidth
Burst defines how much data is sent at once
router(config-pmap-c)#
bandwidth {bw-kbps / remaining percent percentage
percent percentage}
Allocates a fixed amount of bandwidth to a class, in kb/s or percent of the configured
(or default) interface bandwidthCBWFQ portion of LLQsystem
Remaining percent allocates a percentage of available bandwidth
Thetwomain commands that are required to configure an LLQsystem are the priority and the
bandwidth commands.
The prioritycommand is used to identify a class as a strict-priority class andallocate
bandwidthto that class, fhe bandwidthcan be specified cither in kilobits per second or
percentage oftheconfigured ordefault interface bandwidth. Traffic exceeding the specified
bandwidth is dropped if congestion exists, fhe burst option defines theamount of datathat can
be sent at once, "fhe default hurst size is based on 200-ms interval and LLQ bandwidth.
The bandwidth command is used lo allocate bandwidth to nonpriority classes. It applies to the
CBWFQ portion ofthe LLQ system, not tothe priority queue. The bandwidth can bespecified
inkilobits persecond or in percent ofthe configured or default interface bandwidth. The
remaining kev word allows the allocation ofa percentage of remaining (nonal located)
bandwidth.
Note Because bandwidth for the priority class is specifiedas a parameter to the priority
command, youcannot alsoconfigure the bandwidth command fora priority class. If youdo
so, it is a configuration violation that would onlyintroduceconfusionin relation to the amount
of bandwidth to allocate.
LLQ Configuration Example
This subtopic provides an LLQ configuration example.
LLQ Configuration Example
class-map voip
match ip precedence 5
1
class-map mission-critical
match ip precedence 3 4
1
class-map transactional
match ip precedence 1 2
1
policy-map Policyl
class voip
priority percent 10
class mission-critical
bandwidth percent 30
random-de t ec t
class transactional
bandwidth percent 20
random-detect
class class-default
fair-queue
randca-detect
This LLQconfiguration forwards the voip class as strict priority traftic and allocates it 10
percent ofthe interface bandwidth (configured using the bandwidth command or default).
The mission-critical class is allocated 30 percent ofthe interface bandwidth. It uses WREDto
randomly droppacketswhen the queueis fillingup to prevent congestion.
The transactional class gets 20 percent ofthe interface bandwidth and also uses WRED.
The class-default matches all other packets that are not classified into the previousclasses, is
scheduled using the WFQ algorithm, and also uses WRED.
Monitoring LLQ
This subtopic explains how to monitor LLQ.
Monitoring LLQ
router*
j show policy-map interlace interface |
Displays packet statistics of all classes that are configured for all
service policies on the specified interface or subinterface
routerll show policy-map interface Pastetharnet 0/0
PantEthernetO/O
Service-policy output: LLQ
Class-map: LLQ (match-any)
0 packets, 0 bytes
5 minute offered rate 0 bps. drop rate 0 bpa
Match: any
Weighted Fair Queueing
Strict Priority
Output Queue! Conversation 264
Bandwidth 1000 {kbpsl Burst 25000 (Bytes)
{pkts matched/bytes matched) 0/0
{total drops/bytes drops) 0/0
Class-map: class-default (match-any)
0 packets, 0 bytes
5 minute offered rate 0 bps. drop rate 0 bps
Match: any
fhe show policy-map interface command displays thepacket statistics ofall classes thatare
configured for all senice policies on the specified interface. The table describes some ofthe
kev fields in the command output.
Parameter
Description
Class-map
Class of trafficbeing displayed. Output is displayed for each
configured class in the policy.
offered rate
Rate, in kb/s. of packets coming in to the class.
drop rate
Rate, in kb/s. at which packets are dropped fromthe class. The
drop rate is calculated bysubtractingthe number of successfully
transmitted packets from the offered rate.
Match
Match criteria that are specified for the class of traffic.
pkts matched/bytes
matched
Numberof packets (also shown inbytes) matchingthis class that
were placed in the queue.
total drops/bytes
drops
Number of packets/bytes that are discarded for this class.
5-140 Implementing Cisco Voice Communications and QoS(CVOICE) vB.O
Calculating Bandwidth for LLQ
This topic describes how to calculate the bandwidth values for LLQ.
Calculating Bandwidth for LLQ
Bandwidth reservation should include overhead:
Bandwidth used for guarantees and policing with LLQ
For proper operation, it should include Layer 3 and Layer 2
overhead, otherwise it guarantees less than expected
Example: setting up priority queue to support 10 x G.729
calls (20-ms packetization) on MLPPP link
G729
Payload
~r
LLQ pnority bandwidth = 10 x (20 bytes + 40 bytes + 13 bytes) x 50 p/s x 8 bits/byte
= 10x29.2 kb/s = 292 kb/s
LLQ pnority bandwidth rounded up to 300 kb/s
Layer 3+
Overhead
Laye-' i
-.'.V? rise .id
router(config-pmap-c)# priority 300
The bandwidth allocation is specified using two commands within the LLQ structure: the
priority and the bandwidth commands. It delines the bandwidth guarantee and policing rate,
while the policing rate is enforced only during congestion. The bandwidth must include the
Layer 2 and Layer 3 overhead ofthe packets that are forwarded within a class.
"fhe figure illustrates how to compute the bandwidth that is required for an LLQ queue that
should ensure timely forwarding of 10 G.729 calls with 20-ms packetization over a Multilink
PPP (MLP) link.
The frame size is 73 bytes: 20 bytes (payload) + 40 bytes (IP/UDP/RTP) + 13 bytes (MLP).
The bandwidth of a single call is 29.2 kb/s. The aggregate of 10calls totals to 292 kb/s and can
be rounded up to 300 kb/s.
Calculating Bandwidth for
Packsbzabon
Period
G711 20 ms
G711 30ms
G.729 20 ms
G 729 30 ms
160 bytes
240 bytes
20 bytes
30 bytes
Packet* per
Second
802 3 Ethernet
802 1Q Ethernet
PPP
Multilink PPP with
Interleaving
Frame Relay
Frame Relay with FRF.12
Example1 10 G.711 calls, 20-ms packetization, FRF 12
G711
Payload
IB
16+4
LLQ pnority bandwidth = 10 x (160 bytes + 40 bytes
= 10x83 2 kb/s = 832 kb/s
LLQ priority bandvyidth rounded up to 840 kb/s
i-Y-',>)x 50 p/s x 8 bits/byte
router(config-pmap-c)# priority 84C
'fhe figure illustrates how to compute the bandwidth that is required for an I.LQqueuethat
should ensure timeK forwarding of 10 G.711 calls with 20-ms packetization over a Frame
Rela\ link that is configured for frame Relay Fragmentation Implementation Agreement
(FRF.12) fragmentation and interleaving.
The frame size is 208 bytes: 160bytes (payload) + 40 bytes(IP/UDP/RTP) + 8 bytes(FRF.12).
Ihe bandwidthof a singlecall is S3.2 kb/s. The aggregate of 10calls totals to 832 kb/s and can
be rounded up to 840 kb/s.
6-142 Implementing Cisco VoiceCommunications and QoS (CVOICE) v8.0
Summary
This topic summarizes the key pointsthat werediscussedin this lesson.
Summary
Congestion occurs at rate mismatch or aggregation points.
Policing rate-limits or remarks above-limit traffic, while shaping
queues above-limit trafficifwithin allowed burst size.
Traffic rate is measured using single- or dual-token buckets with
the same or different rates.
Dual-rate policing defines two informationrates: committed and
peak, with respective burst sizes (Be and Be).
Class-based policing configuration may specify up to three actions:
conform (below CIR), exceed (above CIR but below PIR), and
violate (above PIR).
Summary (Cont.)
Class-based shaping delays above-threshold packets and does
not re-mark them.
Shaping is configured either to average rate (conservative) or to
peak rate (aggressive).
Low-latency queuing defines a high priorityqueue within a
CBWFQ-based queuing system.
LLQ system is configured using the bandwidth command to
allocate a bandwidth to traffic class and priority command to
declare one class as priority class.
Bandwidth reservation for LLQ must include the Layer 2 and Layer
3 overhead ofthe VoIP calls.
6-144 Implementing CiscoVoice Communications and QoS (CVOICE) vB.O 2010 CiscoSystems, Inc
***
Lesson 51
Understanding Cisco AutoQoS
Overview
Cisco AutoQoS represents two technologies, Cisco AutoQoS VoIP and Cisco AutoQoS for the
Enterprise These technologies simplify network administration challenges, reducing quality of
service (QoS) complexity, deployment time, and cost in enterprise networks.
Cisco AutoQoS VoIP incorporates value-added intelligence in Cisco IOS Software and Cisco
Catalvst software lo provision and manage large-scale QoS deployments. It provides QoS
provistoning for individual routers and switches, simplifying deployment and reducing human
error Cisco AutoQoS VoIP offers straightforward capabilities to automate VoIP deployments
for customers who want to deploy IP telephony but who lack the expertise and staffing to plan
and deploy IP QoS and IPservices.
Cisco AutoQoS for the Enterprise is aprocess in which two intelligent mechanisms are
deployed to detect voice, video, and data traffic in Cisco networks. The mechanisms generate
best-practice QoS policies and apply those policies to WAN interfaces.
This lesson explores the capabilities, requirements, and configuration of Cisco AutoQoS VoIP
andCiscoAutoQoS for theEnterprise.
Objectives
Upon completing this lesson, you will be able to describe Cisco AutoQoS and its operations.
This ability includes being able to meet these objectives:
Explain how Cisco AutoQoS VoIP is used to implement QoS policy and identify the router
and switch platforms on which Cisco AutoQoS VoIP can be used
Describe tlie prerequisites for Cisco AutoQoS VoIP and its configuration using the CLI
Explain how to examine and monitor anetwork configuration after Cisco AutoQoS has
been enabled
Identify the QoS technologies that are automatically implemented on the network using
Cisco AutoQoS VoIP
Explain how Cisco AutoQoS for the Enterprise is used to implement QoS policy
Describe how Cisco AutoQoS for the Enterprise is configured on arouter using the CLI
Describe how to examine and monitor anetwork configuration after Cisco AutoQoS for the
Enterprise has been enabled
Cisco AutoQoS VoIP
'his topic describes hou the Cisco AutoQoS VoIP feature is used to implement QoS policy
6-146
Cisco AutoQoS VoIP
One command per interface to enable and configure QoS
Fine-tuning possible after automatic settings applied
*Available on Cisco IOS routers and switches
" '.yVAN _.' 125 Remote Sues
*-? "
Cisco AutoQoS prov ides the ability lo deploy QoS features for converged IP telephony and
data networks fast and efficiently, by simplifying and automating the Modular QoS command
iteinterface (CLI). or MQC. policy. Cisco AutoQoS generates traffic classes and policy map
CLI templates. \\hen Cisco AutoQoS is configured at the interface or permanent virtual circuit
(I VC). the traffic receives the required QoS treatment automatically. There is no need for in-
depth know ledge ol the underlying technologies, service policies, and link efficiency
mechanisms. Cisco AutoQoS implements best practice for VoIP transport.
Cisco AutoQoS canbe beneficial in these scenarios:
Small- to medium-sized businesses that must deploy IP telephony quickly but lack the
experience andstaffing to plananddeploy IPQoS services
Large enterprises that need to deploy Cisco telephony solutions on alarge scale while
reducing the costs, complexity, and time frame for deployment, and ensuring that the
appropriate QoS for voice applications is being setin a consistent fashion
International enterprises or sen ice providers requiring QoS for VoIP in different regions of
the world where little expertise exists and where provisioning QoS remotely and across
different time /ones is difficult
Service prov iders requiring atemplate-driven approach for delivering managed services
and QoS for voice traffic to manv customer premises devices
mm
Cisco AutoQoS VoIP Functions
This subtopic describes the functions of Cisco AutoQoS VoIP.
Cisco AutoQoS VoIP Functions
Application classification
- Automatically discovers applications
Policy generation
- Automatically generates initial and
ongoing QoS policies
Configuration
- Provides multidevice automation
for QoS
Monitoring and reporting
- Automatic alerts and summary
reports
Consistency
- Interoperability among QoS features
LAN. MAN, and WAN
Cisco AutoQoS VoIP simplifies and shortens the QoS deployment cycle infive major aspects:
Application classification: CiscoAutoQoS leverages intelligent classification on routers
using Cisco Network-Based Application Recognition (NBAR) toprovide deep andstateful
packet inspection. Cisco AutoQoS uses Cisco Discovery Protocol forvoice packets to
ensurethat the deviceattached to the LAN is reallyan IP phone.
Policy generation: CiscoAutoQoS evaluates thenetwork environment andgenerates an
initial policy. Cisco AutoQoS automatically determines WAN settings for fragmentation,
compression, andencapsulation, eliminating theneedto understand QoStheory.
Configuration: With onecommand, Cisco AutoQoS configures the port toprioritize voice
traffic without affecting other network traffic, while still offering theflexibility toadjust
QoSsettingsfor uniquenetworkrequirements. CiscoAutoQoS-generated router andswitch
configurations are customizable usingthe standardCiscoIOSCLI.
Monitoring and reporting: CiscoAutoQoS provides visibility intothe classesof service
that aredeployed via system logging and Simple Network Management Protocol (SNMP)
traps, with notification of abnormal events (thatis, VoIP packet drops).
Consistency: When deploying QoS configurations using Cisco AutoQoS, configurations
that are generated are consistentamongrouter and switchplatforms. This level of
consistency ensures seamless QoS operation and interoperability within thenetwork.
Cisco AutoQoS VoIP Router Platforms
Thistopicdescribes the router platfomis on whichCiscoAutoQoS VoIPcan be used.
Cisco AutoQoS VoJP Router Platfc
Cisco 1800. 2800, 2900. 3800, 3900. and 7200 Series
Routers support Cisco AutoQoS.
VoiceQoS implementation without extensive knowledge of
service policies.
Reduction of the time required to deploy QoS features for
converged IP telephony and data networks.
Cisco AutoQoS lends itself to tuning ofall generated
parameters and configurations.
Support for Cisco AutoQoS includes the Cisco 1800. 2800. 2900. 3800. 3900. and 7200 Series
Routers.
The CiscoAutoQoS VoIPfeature is supported on the following interfaces and PVCs:
Serial interfaces with PPPor lligh-Leve! Data Link Control (HDLC)
Frame Relav data-linkconnection identifiers (DLCTs)point-to-point subinterfaces only
CiscoAutoQoS does not support FrameRelay multipoint interfaces
ATM PVCs
Cisco AutoQoS VoIP is supported on low-speed ATM PVCs on point-to-point
subinterfaces only (link bandwidth less than 768 kb/s)
Cisco AutoQoS VoIP is fully supported onhigh-speed AIM PVCs (link bandwidth
greater than 768 kb/s)
Cisco AutoQoS VoIP Switch Platforms
This topic describes the switch platforms on which Cisco AutoQoS VoIP can be used.
Cisco AutoQoS VoIP Switch Platforms
All current Cisco Catalyst switches support Cisco AutoQoS VoIP.
. voice QoS requirements can be fulfilled without extensive
knowledge of:
- Trust boundary
- CoS-to-DSCP mappings
- WRRand PQ scheduling parameters
. Reduce time required to deploy QoS features for converged IP
telephony anddata networks.
Generated parameters and configurations are user-tunable.
Support for Cisco AutoQoS VoIP includes all current Cisco Catalyst switches.
You can implement the voice QoS requirements without extensive knowledge of Ac trust
boundarv theorv class of service-to-differentiated services code point (CoS-to-DSLf)
rPP.ngs. and weighted round robin (WRR) and priority queuing (PQ) implementation on the
given switch platform.
Once the automatic policy is applied, it can be modified to match the needs of aspecific
environment.
Quality ot Service 6-149
6-150
Cisco AutoQoS VoIP Switch
(Cont.)
- Single command atthe interface level configures interface
and global QoS:
- Support for Cisco IP phone and Cisco IP Softphone
Trust boundary disabled when IP phone ismoved
Buffer allocation and egress queuing dependent on
interface type (Gigabit Ethernet or Fast Ethernet)
Supported on static, dynamic-access, voice VLAN access
and trunk ports
Cisco Discovery Protocol must be enabled for Cisco AutoQoS
to function properly
Ioconfigure the QoS settings and the trusted boundary feature on the Cisco IP phone Cisco
Discover) Protocol version 2or later must be enabled on the port. Ifthe trusted boundary
feature is enabled, asyslog warning message displays ifCisco Diseoverv Protocol is not"
enabled orifCisco Diseoverv Protocol is running version 1.
Cisco Discov erv Protocol needs to be enabled only for the ciscoipphone QoS configuration
Cisco Discovery Protocol does not affect the other components ofthe automatic QoS features
VV hen the ciscoipphone keyword with the port-specific automatic QoS feature is used a
warning displavs ifthe port does not have Cisco Discovery Protocol enabled.
When executing the port-specific automatic QoS command with the ciscoipphone keyword
without the trust option, the trust-device feature is enabled. The trust-device feature is
dependent onCisco Discovery Protocol.
Implementing Cisco Voice Communications and QoS (CVOICE] v8 0
Configuring Cisco AutoQoS VoIP
This topic describes the prerequisites for using Cisco AutoQoS VoIP and how to configure
Cisco AutoQoS VoIP on a network using the CLI.
Cisco AutoQoS VoIP Prerequisites
* Cisco Express Forwarding must be enabled.
No output service policy can exist on the interface.
- Otherwise, error returned with an explanation
There is automatic bandwidth evaluation.
- Low-speed if less than or equal to 768 kb/s
IP address mandatory
High-speed if greater than 768 kb/s
IP address optional
- Requires that correct bandwidth is set on interfaces
or subinterfaces using the bandwidth command.
Before configuring Cisco AutoQoS VoIP, you must meet these prerequisites:
Cisco Fxpress Forwarding must be enabled at the interface or subinterface. Cisco AutoQoS
VoIP uses NBAR to identifyvarious applications and traffic types, and Cisco Express
Forwarding is a prerequisite for NBAR.
No QoS policies (service policies) can be attached to the interface. Cisco AutoQoS VoIP
cannot be configured if a QoS policy exists.
Cisco AutoQoS VoIP classifies links as either low-speed or high-speed depending upon the
link bandwidth. The default bandwidth of a serial interface, if the bandwidth command is
not configured, is 1.544 Mb/s. Therefore, correct bandwidth must be specified on the
interface or subinterface where Cisco AutoQoS VoIP is enabled.
If die interface or subinterface has a link speed of 768 kb/s or lower, an IP address
must be configured on the interface or subinterface using the ip address command.
By default. Cisco AutoQoS VoIP will enable Multilink PPP (MLP) and copy the
configured IP address to the multilink bundle interface.
In addition to the CiscoAutoQoS VoIPprerequisites, the following are recommendations and
requirements when configuring Cisco AutoQoS VoIP:
The Cisco AutoQoS VoIP feature is supported only on PVCs and these interfaces:
ATM PVCs
Serial interfaces with PPP or HDLC
Frame Relay DLCIs (point-to-point subinterfaces only) (Cisco AutoQoS VoIP does
not support Frame Relay multipoint interfaces.)
The configuration template that is generated by configuring Cisco AutoQoS VoIP on an
interface or PVCcan be tuned manually, via CLI configuration, if desired.
Cisco AutoQoS VoIP cannot be configured if a QoS service policy is already configured
and attached to the interface or PVC.
MLP is configured automatically for a serial interface with a low-speed link, fhe serial
interface must have an 1Paddress, which is removed and put on the MLP bundle. Cisco
AutoQoS VoIP must also be configured on the other side ofthe link.
The no auto qos voip command removes Cisco AutoQoS VoIP. Ilowever, if the interface
upon which CiscoAutoQoS VoIP configuration was generated is deletedwithout
configuring the no auto qos >oip command. CiscoAutoQoS VoIPwill not be completely
removed from the configuration.
Cisco AutoQoS SNMPtraps are only deli\ered when an SNMPserver is used in
conjunction with Cisco AutoQoS VoIP,
The SNMP commttnitv string AutoQoS should have write permissions.
If the device is reloaded with the saved configuration aller configuring Cisco AuloQoS
VoiPand savingthe configuration to NVRAM, some warning messages may be generated
bv Remote Monitoring(RMON) threshold commands. These warning messages can be
ignored. (Toavoid further warning messages, savethe configuration to NVRAM again
without making an\ changes lo the QoS configuration.)
6-152 Implementing Cisco Voice Communications andQoS (CVOICE) v8.0 2010Cisco Systems, Inc
mm
Configuring Cisco AutoQoS VoIP: Routers
This subtopic describes how to configure Cisco AuloQoS VoIP on Cisco IPS routers
Configuring Cisco AutoQoS VoIP: Routers
router(config-if)*
router(config-fr-dlci)#
auto qos voip [trust] [fr-atm]
Configures the CiscoAutoQoS VoIP feature
Untrusted mode by default
trust: Indicates that the DSCP markings ofa packetare trusted
{relied on)for classification ofthe voice traffic
fr-atm: Forlow-speed Frame Relay DLCIs interconnected with
ATM PVCs inthe same network, the fr-atm keyword must be
explicitly configured in the auto qos voip command tocorrectly
configureCiscoAutoQoSVoIP.
To configure the Cisco AutoQoS VoIP feature on an interface, use the auto qos voip command
in interface configuration mode or Frame Relay DLCI configuration mode. To remove the
Cisco AutoQoS VoIP feature from an interface, use the no form ofthe auto qos voip
command. The trust keyword indicates that the DSCP markings ofapacket are trusted for
classification ofthe voice traffic. Ifthe optional trust keyword is not specified, the voice traffic
is classified using NBAR, and the packets are marked with the appropriate DSCP value. The fr-
atm keyword enables Cisco AutoQoS VoIP on Frame Relay-to-ATM interworking links.
The bandwidth ofthe serial interface isused to determine the link speed atthe time the feature
is configured. Cisco AutoQoS VoIP does not respond to changes made to bandwidth after the
feature is configured. Ifthe bandwidth is later changed, the Cisco AutoQoS VoIP feature will
not update its policy. To force the Cisco AutoQoS VoIP feature to use an updated bandwidth,
use the no auto qos voip command to remove the Cisco AutoQoS VoIP feature and then
reconfigure the feature.
Configuring Cisco AutoQoS VoIP: Switches
This subtopic describes how to configure Cisco AutoQoS VoIP on Cisco IOS switches.
Configuring Cisco AutoQoS VoIP: Switches
awitch(config-if)#
auto qos voip {cisco-phone | trust}
ConfiguresCiscoAutoQoSVoIP on a switchinterface
Untrusted by default
trust opton specifies that the interface is connected to a trustedswitch
or router and theVoIP classification inthe ingress packetis trusted
Typically configured on upstream interfaces
cisco-phone option.
Automatically enablesthetrusted boundary feature, which uses the
Cisco Discovery Protocol todetect thepresence a Cisco IP phone
If interface isconnected to a Cisco IP phone, QoS labels ofincoming
packets are trusted onlywhen a phone is detected
When the Cisco AutoQoS VoIP feature is enabled on the first switch interface. QoS is globally
enabled (mis qos global configuration command).
When the auto qos voip trust interface configuration command is entered, the ingress
classification on the interface is set to tmst the CoS QoS label that is received in the packet, and
the egress queues on the interface are reconfigured. QoS labels in ingress packets are trusted.
Ihis option is used to identify a port as connected to a trusted switch or router.
When the auto qos voip cisco-phone interface configuration command isentered, the trusted
boundary feature is enabled. The trusted boundary feature uses the Cisco Discovery Protocol to
detect the presence orabsence ofaCisco IP phone. When aCisco IP phone is delected, the
ingress classification on the interface is set to tmst the QoS label that is received in the packet.
When a Cisco IP phone is absent, the ingress classification is sel tonot trust the QoS label in
the packet. The egress queues on the interface arc also reconfigured. This command extends the
trust boundary- if an IPphone is detected.
Monitoring Cisco AutoQoS VoIP
This topic describes how to monitor Cisco AutoQoS VoIP.
Monitoring Cisco AutoQoS VoIP: Routers
router>
show auto qos [interface interface type]
Displays the interface configurations, policy maps, class
maps, and ACLs created automatically
All components can be tuned and examined individually:
- show class-map
- showpoHcy-map
router> show auto qos interface SerialO/1/1
1
interface SerialO/1/1
service-policy output AutoQoS-Poliey-DnTrust
When the auto qos voip command is used to configure the Cisco AutoQoS VoIP feature,
configurations are generated for each interface or PVC. These configurations are then used to
create the interface configurations, policy maps, class maps, and ACLs.
The show auto qos interface command can be used with Frame Relay DIXTsand ATM PVCs.
When the interface keyword is used along with the correspondinginterface type argument, the
show auto qos interface [interfacetype] command displays the configurations that are created
by the Cisco AutoQoS VoIP feature on the specified interface.
When the interface keyword is used but an interface type is not specified, the show auto qos
interface command displays the configurations that are created by the Cisco AutoQoS VoIP
feature on all the interfaces or PVCs on which the Cisco AutoQoS VoIP feature is enabled.
>2010CiscoSystems, Inc.
Monitoring AutoQoS VoIP; Routers (Cont)
Sample output for low-speed Frame Relay interface
router* show auto qoa interface srs/1.1
Serial6/1.1; DLCI 100 -
interface Serial6/1
frame-relay traffic-shaping
interface Serial6/l.l point-to-point
class AutoQoS-VoIP-PR-Serial6/I-100
frame-relay ip rtp header-compression
map-class frame-relay AutoQS-VoIP-FR-Seria16/1-100
frame-relay be 5120
frame-relay be 0
frame-relay mincir 512000
service-policy output AutoQoS-Policy-UnTrust
frame-relay fragment 54 0
The shon auto qos command prov ides more information than the show auto qos interface
command, which onlv shows the service policy that is applied lo an interlace. The show auto
qos command can be usedto verifv the contents ofthe interface configurations, policy maps,
class maps, and access control lists (ACLs).
Theexample inthe figure presents theshow autoqos command ina scenario in which the
AutoQoS VoIPfeature was configured at a low-speed frame Relay interface.
Monitoring AutoQoS VoIP: Switches
This subtopic describes how to monitor Cisco AutoQoS VoIP on Cisco Catalyst switches.
Monitoring AutoQoS VoIP: Switches
show auto qos [interface interface-id]
Displays the initial Cisco AutoQoS settings
Does not display any user changes to the autoconfiguration
- Added as a result of tuning
soitcbl show auto qos
initial configuration appli sd by Aut jQoS:
wrr queue bandwidth 2 1 80 0
no vrr-queue cos-map
wrr queue coo 10 12 4
vrr que e coa 3 3 6 7
wrr queue cos 4 5
mis qos map cos-dscp S 16 26 32 46 48 56
int arfa e FaatEthemetO/3
mis qoa trust device c laco- phone
mis qos trust cos
To display the initial Cisco AutoQoS VoIP configuration, use the show auto qos [interface
[interface-id]] privileged EXEC command. To display any user changes to that configuration,
use the show running-config privileged EXEC command. You can compare the show auto
qos and the show running-config command output lo identify the user-defined QoS settings.
>2010 Cisco Systems. Inc. Quality of Service 6-J57
Automation with Cisco AutoQoS
This topic describes several ofthe QoS technologies that are automatically implemented on the
network when using Cisco AutoQoS.
Automation with Cisco AutoQoS
DiffServ
Functfon
Cisco IOS Software
QoS Feature
Applicability Description ';;;.
Classification NBAR DSCP. ACL Trust op Bon:
only
DSCP-based
Classification of VoIP; when trust
option used, classification based
on DSCP only
Marking Class-based marming Only when
untrusted
Set Layer 2 and Layer 3
attributes when trust option not
Congestion
Management
Percentage-based
LLQ
Always Provide EF treatment to voice
and best-effort treatment to data
Shaping FRTS Frame Relay Shape to CIR to smooth traffic to
configured rate
Compression RTP header
compression
Stow links
(<768 kb/s)
Increases throughput; enabled
only on slow links
LFI MLP with interleaving
or FRF.12
Slow PPP and
Frame Relay
iinks
Reduces voice jitter; enabled
only on slow links
Cisco AutoQoS performs these functions in a WAN:
Automatic classification of Real-1 ime Iransport Protocol (RTP) payload and VoIP control
packets (H.323. H.225 Unicast. Skinny Client Control Protocol [SCCP]. Session Initiation
Protocol [SIP], and Media Gateway Control Protocol [MGCPJ) using NBAR and ACLs.
Thetrust option enables classification using DSCP values.
Marking is performed when the trust option is not used.
Builds service policies fbr VoIP traffic that arc based on Cisco MQC.
Provisions low latencv queuing (I.LQ) for VoIP bearer and bandwidth guarantees fbr
control traffic.
Enables Frame Rela; traffic shaping that adheres to Cisco best practices, where required.
Enables link eftlciencj mechanisms such as link fragmentation and interleaving (I.FI) and
compressed Real-TimeTransport Protocol (cRTP) on slowlinks.
Provides SNMP and syslog alerts for VoIP packet drops.
isco AutoQoS performs these functions in a LAN:
Enforces the trust boundary on switch access ports, and uplinks and downlinks.
Enables Cisco Catalyst strict PQ(also known as expedited queuing) with WRR scheduling
for voice and data traffic, where appropriate.
Configures queue admission criteria (maps CoS values in incoming packets to the
appropriate queues).
Modifies queue sizes and weights where required.
6-158 Implementing Cisco Voice Communications and QoS (CVOICE) vB.O 2010 Cisco Systems Inc
Cisco AutoQoS for the Enterprise
This topicdescribes howCiscoAutoQoS for the Enterprise is usedto implement QoSpolicy.
Cisco AutoQoS for the Enterprise
WAN QoS policy
automation tool
Supported on
routers only
Medium-to-large
campus with large
remote sites
Two commands entered
on the WAN interface
Classification for up to
10 traffic classes
Too granular for
many environments
- Fine-tuning required
to merge traffic
classes
The Cisco AutoQoS for the Enterprise feature automates the deployment of QoS policies in a
general business environment. The policies that are deployed by Cisco AutoQoS for the
Enterprise are not solely focused on VoIP convergence, but also on convergence of voice,
video, and data applications. Cisco AutoQoS for the Enterprise is generally deployed in midsize
companies and branch offices of larger companies. It is used to provide best-practice QoS
policy generation for voice as well as to provide for classification of up to 10traffic classes.
Existing QoS policies may be present during the first configuration phase of Cisco AutoQoS
for the Enterprise: that is, during the autodiscovery (data collection) phase. However, any
existing QoS policies must be removed before the Cisco AutoQoS-generatedpolicies are
applied during the second configuration phase of this feature.
Cisco AutoQoS for the Enterprise classifies the traffic using a 10-class model, which is loo
wide for many environments, especially with low-speed interfaces. In such cases, manual
tuning should be performed to merge the classes into fewer traffic types.
)2010CiscoSystems, Inc.
Cisco AutoQoS for Enterprise (Cont
Two components
- Autodiscovery
- Cisco AutoQoS policy
Autodscovery j CfctfoAutoOoS Pottey-
Network- Based
Application Recognition GscoAutoQoS
(NSAR)discovers Class maps
applications and Match statements
protocol tvpes
Offered tut
rate (average
and peak)
Mnmum bandwidth
to class queues,
scheduling, and WRED
Cisco AutoQoS for the Enterpriseis perfomied in the following two phases:
Autodiscovery: fhe autodiseoven process detects applications and protocol types. The
autodiscovery phasegivesthe discovery periodenough timeto gather trafficstatistics
beforeapply ingCiscoAutoQoS. Thetime for gathering trafficstatisticsbefore applying
CiscoAutoQoS depends upon the actual highs and lowsof trafficpatterns of a network.
Provisioning: Cisco AutoQoS policy is the provisioning stage, inwhichthe appropriate
MQC configuration is applied to the router. It includes the recommended QoSpolicythat is
generated and installed by the AuloQoS macros, basedon the data that is gathered by the
autodiscoven process.
The figure shows the 10 traffic classes that canbe classified byCisco AutoQoS for the
Enterprise. They may ha\e tobeaggregated intofewer classes during the manual tuning of the
automatic Cisco AutoQoS configuration.
6-160 Implementing CiscoVoice Communications and OoS (CVOICE) v80
Configuring Cisco AutoQoS for the Enterprise
This topic describes how to configure Cisco AutoQoS for the Enterprise^
Configuring Cisco AutoQoS for the
Enterprise
router(coafig-if)#
router(config-r-dlci)#
auto discovery qoe [trust] =1
Enables phaseone (autodiscovery)
- Trust option trustsingress marking
Requires Cisco Express Forwarding (prerequisite for NBAR)
Existing QoS policy may exist during discovery
Traffic statistics are gathered by NBAR
Interface speed determines theconfiguration
- Stop and restart autodiscovery if bandwidth is changed
Discovery period:
- Must provide arepresentative sampling ofthe volume and
type ofvoice, video, and data
- Depends on the business cycle of organization
The auto discovery qos [trust] command starts the discovery phase *^*?^
subinterface. or Frame Relay DLCI. The discovery stage must be completed before the Cisco
AutoQoS policy can be applied to the router interface. When the discovery is complete, the
results can be viewed using the show auto discovery qos (interface) command. The router wtll
not accept the auto qos command until statistics have been gathered.
The trust keyword is used to make the router rely on the markings in ingress packets.
Cisco AutoQoS for the Enterprise has the same prerequisites as Cisco AutoQoS VoIP, such as
the following:
. Cisco AutoQoS uses NBAR to identify various applications and traflic types, and Cisco
Express Forwarding is aprerequisite for NBAR.
. Although an existing QoS policy is allowed during the discovery stage, it must be removed
beforeCiscoAutoQoS commitsits policy.
Cisco AutoQoS classifies links as either low-speed or high-speed, depending upon the link
bandwidth The correct bandwidth configuration is necessary to apply the correct policy.
When the bandwidth is changed during the discovery phase, the autodiscovery must be
stopped (using the no auto discovery qos command) and restarted (using the auto
discovery qos command) to take the new setting into account.
The discovery period should cover the complete business cycle of an enterprise and typically
ranges between 2 and7 days.
6-162
Configuring Cisco AutoQoS forth*
Enterprise (Cont.)
router(config-if1#
router(config-fr-dlci) #
|auto qos
' Enables second phase for Cisco AutoQoS for Enterprise
Requires that statistics have been gathered from phase one
Generates QoS policy based on NBAR statistics and installs
the service policies inthe WAN interface
No service policy canexist onthe interface
Must be applied on both ends ofslow links
Otherwise compression and LFI disrupts connectivity
J
The auto qos command commits the automatically generated QoS policy to the router This
command ,s accepted after the discovery statistics have been gathered. No QoS service poliev
can exist on the interface. ' -
On slou frame Relay links (less than 768 kb/s). the Cisco AuloQoS poliev provisions the
frame Relay fragmentation Implementation Agreement (FRE.I2). The fragmentation breaks
connectivity over the link, unless it is also configured on the other side ofthe PVC.
Monitoring Cisco AutoQoS for the Enterprise
Thistopic describes howtomonitor Cisco AutoQoS for the Enterprise.
Monitoring Cisco AutoQoS for the
Enterprise: Phase 1
router#
show auto discovery qos
Displays autodiscovery results; gathered by NBAR
routert show auto discovery qos
AutoQoS Discovery enabled for applications
Discovery up tine: 2 days, 55 minutes
AutoQoS Class information;
Class VoIP:
Recommended Minimum Bandwidth: 517 Kbps/50% (PeakRate)
Detected applications and data:
Application/ AverageRate PeakRate Total
Protocol (kbps/*> tkbps/%) (bytasl
rtp audio 7 6/7 517/50 703104
Class Interactive Video:
Recommended Minimum Bandwidth: 24 Kbp /2% (AverageRate)
Detected applications and data:
Application/ AvarageRate PeakRate Total
Protocol (>tbps/%) (kbpo/%) (bytes)
rtp video 24/2 5337/52 704574
Theautodiscovery qoscommand displays theresults ofthe discovery phase. The figure shows
an example of NBAR recognizing traffic classes that have been detected on a WAN interface.
The information includes the lengthofthe discovery periodandthe detectedtraffic classes,
with their throughput rates.
Monitoring Cisco AutoQoS for the Enterprise: Phase 2
This topic describes hou tomonitor thecommit phase of Cisco AutoQoS fbrthe Enterprise.
nitoring Cisco
Enterprise: Phase
router#
show auto qos [interface interface type]
Displays the interface MQC configurations
router* show auto qos interface Serial2/l.l
class-map mate): -any AutoQoS voice -Se2/1 1
match protocc 1 rtp audio
class-map rnatct -any AutoQoS Inter -video Se2/l.l
match protocc 1 rtp video
class-map matct -any AutoQoS yigna ling-S 32/1.1
Batch protocc 1 Sip
match protocc 1 rtcp
policy- map Ante QoS-Policy-S B2A.1
claas AutoQoS Voice-Se2/1. 1
prior ity per; ent 3 3
set d scp e
class AutoQOS Inter-Vic1eo- S62/1. 1
bandwidtb ten a in Lug pares Qt 1
set d scp af4
The show auto qos command displays the initial policy that iscommitted by Cisco AuloQoS
fbr the Enterprise. The interface keyword allows amore selective display ofinterface-specific
configuration.
The figure shows asample QoS policx configuration that isgenerated by Cisco AutoQoS
templates, based on NBAR statistics that arc gathered during (he data discovery phase. Hie
configuration is built in modular way. based on MQC building blocks (class-maps, policy-
maps, sen ice policy). The committed rules can be tuned, 'funing is often performed to reduce
the number of traffic classes from 10 to a lower number.
6-164 Implementing Cisco Voice Communications andQoS(CVOICE! v80
Summary
This topic summarizes the key points that were discussed in this lessorn^
Summary
. Cisco AutoQoS VoIP automates VoIP-related QoS provisioning
and is supported on Cisco routers and switches.
Cisco AutoQoS VoIP is enabled with asingle interface command,
and candeclare the interface as trusted.
. The show auto qos command is used to verify the interface
configurations, policy maps, class maps, and ACLs.
Cisco AutoQoS automates the provisioning of classification,
marking, congestion management, shaping, and link efficiency
mechanisms.
Cisco AutoQoS for the Enterprise is an automated QoS
provisioning tool for router WAN links.
Cisco AutoQos for the Enterprise consists of two phases:
autodiscovery and policy provisioning.
Both phases of Cisco AutoQoS for the Enterprise can be verified.
6-166 Implementing Cisco Voice Communications and QoS (CVOICE] vS.O 201Q Cisco Systems, Inc
Module Summary
Module Summary
QoS ensures timely delivery for VoIP packets and can be
implemented using the CLI, MQC, AutoQoS, and QPM.
The DiffServ model strikes a balance between the granularity
of QoS guarantees and ease of deployment.
Classification and marking enable DiffServ PHB operations
by coloring the traffic at the network edge. Link efficiency
mechanisms, such as compression, fragmentation, and
interleaving are mandatory on slow links.
Policing, shaping, and queuing are deployed to manage
congestion in the network.
Two Cisco AutoQoS types, Cisco AutoQoS VoIP and Cisco
AutoQoS for the Enterprise, provision state-of-the-art QoS
methods.
This module describes quality of service (QoS) implementationin VoIP networks. It identifies
the requirements for appropriate transmission service, suchas bounded one-waydelay (150
ms). bandwidth requirements that are specifiedby the selectedcodec, acceptable packet loss (1
percent), and jitter (30 ms). QoS mechanisms can be deployed using four methods: command-
line interface (CI.I), Modular QoSCLI (MQC), CiscoAutoQoS, and QoS Policy Manager
(QPM). The three majorQoSmodelsincludebest-effort, Integrated Services (IntServ). and
differentiated services(DiffServ). DiffServ is the preferredchoicewhendeploying QoSin
modern networks. DiffServ requires thatmostoperations, suchas classification andmarking,
are performed at the network edge, while the devices in the middle ofthe network are
responsible for providing the appropriate per-hop behavior (PHB). Slow links, with
transmission speeds lessthan 768kb/s, require special attention inordertoofferadequate VoIP
transport. They mustuselinkefficiency mechanisms, such as compression, fragmentation, and
interleaving to ensure a timely forwarding of VoIP packets. Policing, shaping, andqueuing
constitute a groupof methodsthat are concerned withmanagingcongestion inthe network.
They focus on dropping, remarking, or queuingof excess packets. CiscoAutoQoS offersthe
easiest wayto deploy QoSin the network and is available intwo forms: CiscoAutoQoS VoIP
and Cisco AutoQoS for the Enterprise.
Module Self-Check
Use the questions here to reviewwhat you learned in this module. The correct answers and
solutions are found in the Module Self-Check Answer K.ey.
Q1) Which term describes the time that it takes to actually transmit a packet on a link ("put
bits on the wire"')? (Source: Introducing QoS)
A) encoding delay
B) processing delay
C) serialization delay
D) transmission delay
Q2) What are the two least expensive solutions for reducing delay on a link? (Choose two.)
(Source: Introducing QoS)
A) Compress data and headers.
B) Drop low-priority packets early.
C) Increase the bandwidth ofthe link.
D) Incorporate advanced queuing technologies.
E) Reduce link MTU size.
Q3) Which two features characterize converged network traffic? (Choose two.) (Source:
Introducing QoS)
A) no low-speed links
B) time-sensitive packets
C) network protocol mix
D) intolerance of brief outages
E) bursty small packet flow
Q4) How much one-way delay can a voice packet tolerate? (Source: Introducing QoS)
A) 15 ms
B) 150 ms
C) 300 ms
D) 200 ms
Q5) RTP helps synchronize real-time transmissions such as voice by time-stamping packets
so that they can be resynchronized at the receiving end. This helps to minimize .
(Source: Introducing QoS)
A) discards
B) splitting
C) tail drop
D) jitter
Q6) Which two statements describe the definition of a QoS policy? (Choose two.) (Source:
Introducing QoS)
A) user-validated
B) networkwide
C) time-based
D) ports that are opened on firewalls
E) different classes of network traffic
>2010Cisco Systems, Inc. Quality ofService 6-169
Q7) How can you fine-tune the QoS implementations that are generated with Cisco
AutoQoS? (Source: Introducing QoS)
A) CLI
B) Modular QoS CU
C) QoSAutoTunc
D) QoS Policy Manager
Q8) Which two advantages does the MQC provide? (Choose two.) (Source: Introducing
QoS)
A) (HT-based
B) ability to apply one policy to multiple inlerfaces
C) separation of classification from policy definition
D) automatic generation of CI.! commands from MQC macros
E) easiest to deploy
Q9) Which QoS implementationmethod offers the shortest implemenlationtime fbr simple
networks? (Source: Introducing QoS)
A) CU
B) MQC
C) AutoQoS
D) AutoTuner
QIO) Which of these models for implementing QoS is the least scalable? (Source:
Introducing QoS)
A) best effort
B) Integrated Services
C) Differentiated Seniccs
D) Quantitative Services
Ql 1) What is the most important advantage of DiffServover olher QoS models? (Source:
Understanding QoS Mechanisms and Models)
A) high scalability
B) manv service levels
C) guaranteed ser\ ice
D) deterministic delays
E) advanced queuing mechanisms
Q12) Services are provided to whichentitiesin the DiffServ model?(Source: Understanding
QoS Mechanisms and Models)
A) frames
B) packets
C) applications
D) classes of traffic
Q\3) How manv bits belongto the DSCP fieldofthe IPheader?(Source: Understanding
A)
~\
B) 4
C) 6
D)
8
6-170 implementing Cisco Voice Communications and QoS (CVOICE) v8 0 2010 Cisco Systems, Inc
Q14) Which PHB is indicated by the DSCP valueof 46 (101110)? (Source: Understanding
A) default PHB
B) selector PHB
C) AF PHB
D) EF PHB
Q15) Which AF class and which drop probability is indicated by the DSCP value of 100100?
(Source: Understanding QoS Mechanisms and Models)
A) AF1 and medium
B) AF4 and medium
C) AK1 and high
D) AF4 and high
Q16) Match each QoS mechanism to its function. (Source: Understanding QoS Mechanisms
and Models)
A) congestion avoidance
B) congestion management
C) classification
D) traffic policing
E) traffic shaping
F) packet header compression
1. drops misbehaving traffic to maintain network integrity
2. improves the bandwidth efficiency of a link
3. controls traffic by delaying bursts
4. discards specific packets based on markings
5. identifies and splits traffic
6. prioritizes, protects, and isolates traffic based on markings
Q17) Which QoS mechanism is used on both input and output interfaces? (Source:
Understanding QoS Mechanisms and Models)
A) classification
B) traffic policing
C) traffic shaping
D) congestion management
Q18) Which three traffic classes are identified in the 11-class Cisco baseline model? (Choose
three.) (Source: Understanding QoS Mechanisms and Models)
A) telepresence
B) streaming video
C) interior gateway protocols
D) scavenger
H) RSVP
F) audio conferencing
G)
video conferencing
)2010Cisco Systems, Inc. Quality of Service 6-171
Q19) Which command would you use to attach a QoS policy to an interface? (Source:
Explaining Classification. Marking, and Link Efficiency Mechanisms)
A) policy-sel-intcrface
B) policy-map
C) policy-interface
D) service-policy
Q20) Classification of packets should occur . (Source: Explaining Classification.
Marking, and Link Efficiency Mechanisms)
A) at the distribution laver
B) anv where in the core ofthe network
C) as close to the source ofthe Iraffic as possible
D) as close to the destination ofthe traffic as possible
Q21) To use a class map. QoS must be referenced by using . (Source: Explaining
Classification. Marking, and Link Efficiency Mechanisms)
A) a route map
B) an access list
C) a poliev map
D) a service map
Q22) What is a requirement for using class-based marking? (Source: Explaining
A) Cisco Express Forwarding must be enabled.
B) Cisco Express Forwarding must be disabled.
C) Cisco Express Forwardingcan only be used on serial inlerfaces.
D) CiscoExpress Forwarding can only be usedon Ethernet interfaces.
Q23) What is the MQCfeature that allows traffic to be classified by a packet subport
number? (Source: Explaining Classification. Marking, and Link Efficiency
Mechanisms)
A) LDPM
B) NBAR
C) service maps
D) service classes
Q24) How do Cisco Unified Communications Manager Express endpoints mark media
traffic?(Source: Explaining Classification. Marking, and Link Efficiency Mechanisms)
A) with IP precedence value of 5
B) with DSCP value of EF
C) with the DSCP value obtained via TF1T (default: LF)
D) with the DSCPvalue obtained via DIICP(default: EF)
E) with the DSCPvalue obtained via Cisco Discovery Protocol (default: EF)
Q25) QoS is enabled on Cisco Catalyst switches bydefault. (Source: Explaining
A) true
B) false
6-172 Implementing Cisco Voice Communications and OoS (CVOICE) v8 0 2010 Cisco Systems, Inc
Q26) Which two of these can be configured as trusted using the trust boundary feature on
Cisco Catalyst switches? (Choose two.) (Source: ExplainingClassification, Marking,
and Link Efficiency Mechanisms)
A) the fact that a legitimate Cisco Unified IP phone is attached to the port
B) the IP precedence value
C) Layer 2 marking in 802.1Q frames
D) Layer 2 marking in 802.3 frames
E) Layer 2 marking in ISL frames
F) the marking set in the first six bits ofthe ToS octet
Q27) Which command will display both the CoS-to-DSCP and DSCP-to-CoS mappings on a
Catalyst switch? (Source: Explaining Classification, Marking, and Link Efficiency
Mechanisms)
A) show mis maps
B) show mis qos maps
C) show mis maps both
D) show qos mis maps both
Q28) Which two of these statements are true, based on the following show output? (Choose
two.) (Source: Explaining Classification, Marking, and Link Efficiency Mechanisms)
wg6rl#show policy-map interface sO/1
Serial0/1
Service-policy output: custl
Class-map: voip (match-all)
300 packets, 202035 bytes
5 minute offered rate 2000 bps, drop rate 0 bps
Match: protocol rtp
Queuing
Strict Priority
Output Queue: Conversation 264
Bandwidth 384 (kbps) Burst 9600 (Bytes)
{pkts matched/bytes matched) 232/139975
(total drops/bytes drops! 15/15795
compress:
header ip rtp
UDP/RTP compression:
Sent: 285 total, 284 compressed,
9086 bytes saved, 176014 bytes sent
1.5 efficiency improvement factor
99% hit ratio, five minute miss rate 0 misses/sec, 0 max
rate 2000 bps
A) Both class-based TCP and RTP header compression are enabled for the VoIP
traffic class.
B) IP payload compression is enabled for the VoIP traffic class.
C) LLQ is enabled for the VoIP traffic class.
D) Class-based RTP header compression is enabled for all RTP traffic.
i 2010CiscoSystems, Inc. Quality of Service 6-173
Q29) Which two of these statements regarding TCP and RTP header compression are
correct? (Choose two.) (Source: Explaining Classification, Marking, and Link
Efficiencv Mechanisms)
A) Hardware-assisted header compression is required to reduce the header
compression delav1.
B) I CP header compression compresses both the IP and TCP headers.
C) RIP header compression compresses the IP. UDP, and RTP headers.
D) TCP and RTP header compression can compress the respective protocol
headers down to 10 bytes.
Q30) Which QoS mechanisms should be deployed on high-speed links? (Source: Explaining
Classification. Marking, and Link Efficiencv Mechanisms)
A) cRTP
B) III
C) class model 5 to 11 classes
D) RSVP
Q31) When configuring Multilink PPPwith interleaving, where is the fragment size
configured? (Source: Explaining Classification, Marking, and Link Efficiency
Mechanisms)
A) Ihe fragment si/e in milliseconds is configured under the physical serial
interfaces with PPPencapsulation and Multilink PPPenabled.
B) Ihe fragment size in milliseconds is configured underthe logical multilink
interface.
C) fhe fragment size in bytes is configured underthe map class.
D) fhe fragment size in bvles is configured within the policymap.
Q32) Which two factors infiuencc the serialization delay?(Choose two.) (Source: Explaining
A) link speed
B) speed oflight in the media
C) router CPl! processing power
D) packet si/e
Q33) 1o ensuregoodvoicequalitv. what is the recommended fragment si/e? (Source:
Explaining Classification. Marking, and Link Efficiency Mechanisms)
A) 80bvtcs per even 64 kb/s ofthe clocking, which will result ina 10-ms
serialization delav
B) 40 bvtes per everv 64 kb/s ofthe clocking, which will result in a 20-ms
serialization delav
C) 20bvtes pereven 64kb/softhe clocking, which will result ina 10-ms
serialization delay
D) 120 bvtes per ever. 64 kb/s of the clocking, which will result in a 150-ms
serialization delav
Q34) Each fragment isa self-sufficient frame, which means that fragmentation should be
configured on therouters that areadjacent to thetraffic source anddestination.
(Source: Explaining Classification. Marking, andLink Efficiency Mechanisms)
A) true
B) false
6-174 Implementing Cisco Voice Communications andOoS (CVOICE) v80 2010Cisco Systems, Inc.
Q35) Tlie show interfaces multilink command output should display which states as open to
indicate proper Multilink PPP operation over an IP interlace? (Source: Explaining
A) LCP, multilink, andIPCP
B) LCP. IPCP, and EEICP
C) LCP and MLP
D) LCP and IPCP
Q36) Which configuration element is used to provision FRE.12 fragmentation? (Source:
Explaining Classification, Marking, and Link Efficiency Mechanisms)
A) class map
B) policy map
C) service policy
D) map class
E) map policy
Q37) Given that the hardware queue is not foil, how will the next packet be serviced by the
software queue? (Source: Managing Congestion and Rate Limiting)
A) Thesoftware queue will be bypassed.
B) Thesoftware queue will enqueue the packet.
C) The software queue will expedite thepacket.
D) The software queue will only meter thepacket.
Q38) Which ofthese is the default dropping scheme for CBWFQ? (Source: Managing
Congestion and Rate Limiting)
A) RED
B) WRED
C) tail drop
D) class-based policing
Q39) In which way does LLQ extend CBWFQ? (Source: Managing Congestion and Rate
Limiting)
A) strict priority scheduling
B) alternate priority scheduling
C) nonpoliced queues for low-latency traffic
D) special voice traffic classification and dispatch
Q40) To which type oftraffic should you limit the use ofthepriority command? (Source:
ManagingCongestion and Rate Limiting)
A) critical data traffic
B) voice traffic
C) bursty traffic
D) video and teleconferencing traffic
>2010 Cisco Systems, Inc. Quality ofService 6-575
Q41) Which ofthese is amajor difference between traffic policing and traffic shaping?
(Source: Managing Congestion and Rate Limiting)
A) Traffic policing drops excess traffic, while traffic shaping delays excess traffic
bv queuing it.
B) Traffic policing is applied only in the outbound direction, while traffic shaping
can be applied to both the inbound and outbound directions.
C) Traffic policing is not available on Catalyst switches, while traffic shaping is
available on Catalyst switches.
D) '['raffle policing requires policing queues to buffer excess traffic, while traffic
shapingdoes not requireany queuesto bufferexcesstraflic.
Q42) Which mathematical model is used by traffic policing mechanisms to meter traffic?
(Source: Managing Congestion and Rate Limiting)
A) token bucket
B) RED
C) FIFO metering
D) Predictor or Stacker
Q43) Which two situations tvpically require the use ofarate-limiting mechanism? (Choose
two.) (Source: Managing Congestion andRate Limiting)
A) Frame Relay connection with speed mismatch at both endpoints
B) aggregation at thedistribution layer
C) service provider providing subrafc access
D) inthe serv ice prov iderhigh-speed backbone, toavoid congestion
E) reducing frame errors on an unstable interface
Q44) When coniiguring single-rate class-based policing, which configuration parameter is
used to enable adual token bucket? (Source: Managing Congestion and Rate Limiting)
A) configuring a violate action
B) configuring an exceed action
C) configuring the PIR in addition to the CIR
D) configuring Be
Q45) What isthe main advantage of using multiaclion policing? (Source: Managing
Congestionand Rate Limiting)
A) to distinguish between exceeding and violating traffic
B) to distinguish between confomiing and exceeding traffic
C) to allow the settingof both Layer2 and Layer3 QoSmarkersat the same time
D) to allow marking ofthe traffic before transmission
Q46) What arethe two configuration options when configuring class-based traffic shaping?
(Choose two.) (Source: Managing Congestion and Rate Limiting)
A) shape average
B) shape peak
C) single or dual token bucket
D) single or multiaction traffic shaping
E) single- or dual-rale traffic shaping
6-176 Implementing Cisco Voice Communications andQoS(CVOICE| v80 2010 Cisco Systems.Inc.
Q47) Which two items are features of class-based policing? (Choose two.) (Source:
Managing Congestion and Rate Limiting)
A) single or dual token bucket
B) single or multiaction policing
C) single- ordual-threshold policing
D) single or dual FIFO queuing
E) single or dual dropthreshold
Q48) 1low should the bandwidth requirement be calculated for the LLQ queue? (Source:
Managing Congestion and Rate Limiting)
A) excluding any Layer 2or Layer 3+ overhead
B) including Layer 3+overhead
C) including Layer 2+overhead
D) 2* (codec bandwidth) *(number ofcalls)
Q49) What is "trusted" when the auto qos voip command is configured with the trust
parameter on routers? (Source: Understanding Cisco AutoQoS)
A) source address
B) MAC address of sender
C) DESkeyword
D) DSCP
Q50) Which ofthese terms is displayed by the show auto qos interface SO/0 router
command? (Source: Understanding Cisco AutoQoS)
A) ACL configuration that isapplied tothe SO/0 interface
B) classmaps configuration
C) policy maps configuration
D) service policy configuration onthe SO/0 interface
Q51) Which command would you use on aCisco IOS Software-based Catalyst switch to
display the configuration ofthe egress queues? (Source: Understanding Cisco
AutoQoS)
A) show mis qos maps
B) show auto qos
C) show auto qos voip
D) show mis qos interface
Q52) Which two mechanisms does Cisco AutoQoS VoIP automatically configure on aWAN
interface? (Choose two.)(Source: Understanding Cisco AutoQoS)
A) enable payload compression
B) provision LLQ
C) markvoicetrafficwith DSCPvalueof EF
D) provision CBWFQ
E) enable LFI where required
)2010 Cisco Systems, Inc. Quality ofService 6-177
Q53) Which two mechanisms docs Cisco AutoQoS VoIP automatically provision on high
speed Frame Relay interfaces? (Choose two.) (Source: Understanding Cisco AutoQoS)
A) LLQ
B) marking
C) shaping
D) cRI'P
E) LFI
Q54) What are two limitations ofCisco AutoQoS for the Enterprise? (Choose two.) (Source:
Understanding Cisco AuloQoS)
A) availability onlv in selected Cisco IOS feature sets
B) no support for Catalyst switches
C) classification into 10 traffic classes
D) required expertise for deployment
E| support for a few interface tvpes
Q55) Order the commands in the most logical sequence for deployment ofCisco AutoQoS
for the Enterprise. (Source: Understanding Cisco AutoQoS)
show auto qos
auto discovery qos
show auto discovery qos
auto qos
6-178 implementingCisco Voice Communicationsand QoS (CVOICE) vS.O 2010 Cisco Systems. Inc.
r*"
Ql]
C
Q2) A.D
Q3)
B,D
Q4) B
Q5)
D
Q6) B,E
Q7) A
Q8) B,C
Q9) C
QIO) D
Qtl) A
Q!2) D
Ql->)
C
QH) D
Q15) B
016) 1-D
2-F
3-t
4-A
5-C
6-B
Q17) B
Q18)
B. D. G
Q1Q) D
Q20) C
Q21) C
Q22) A
Q23) B
Q24] C
Q25) B
Q26) C, F
Q27) B
Q28) C,D
Q29) B,C
Q30) C
Q31)
B
Q32)
A.D
Q33) A
Q34) B
Q35) A
Q36) D
Q37) A
Q38i C
Q39) A
0401 B
Q4I) A
042) A
Q43) A. (_
044i A
04.M C
046) A. B
04-) A. B
0-18) C
049) D
050) D
0?0 B
QS2) B. 1
053) A. C
054) B.C
1 auto disco\er> qos
2 show auto discoveryqos
3. auto qos
I. show auto qos
Implementing Cisco Voice Communications andQoS (CVOICE) vS.O
Table of Contents
Lab Guide .
Overview
Outline ^
Pullout Pod Layout and IP Addressing 2
Description of Simulated PSTN ^
Simulated PSTN Phone j
Calls Inbound to Simulated PSTN J
Calls Outbound from Simulated PSTN 8
Lab 1-1: Configuring Voice Ports 1"[
Activity Objective []
Visual Objective ''
Required Resources **
Command List ^
Job Aids 13
Task 1: Configure DHCP Servers toSupport Autoregistration 14
Task 2: Autoregister Cisco Unified IP Phones 15
Task 3: Configure PRI Interfaces 16
Task 4: Enable HQ to PSTN Calls 16
Task 5: Enable PSTN to HQ Calls 18
Task 6: Enable BRto PSTNCalls 18
Task 7: Enable PSTN-to-BR Calls 19
Lab 1-2: Configuring DSPs 20
Activity Objective 20
Visual Objective 20
Required Resources 20
Command List 21
Job Aids 21
Task 1: Operate DSPs in Default Codec Complexity Mode Flex 22
Task 2: Operate DSPs inMedium Codec Complexity Mode 23
Lab 2-1: Configuring VoIP Call Legs 25
Visual Objective 25
Command List 26
Job Aids 27
Task 1:Configure Basic VoIP onthe HQ Gateway 27
Task2: Configure Codecon the HQ Gateway 28
Task3: Configure Asymmetric CodecNegotiation 28
Task4: Configure Symmetric CodecNegotiation and Examine Fast Start 29
Task5: Configure Slow Startandthe Interface Bind Feature 30
Task6: Configure SIPSignaling 30
Lab3-1: Configuring CiscoUnified Communications ManagerExpress to Support Endpoints 32
Visual Objective 32
Command List 33
Job Aids 34
Task 1: Delete Existing SCCP Endpoints on HQGateway 34
Task 2: Configure CiscoUnified Communications ManagerExpress System Parameters for
SCCP Endpoints 34
Task 3: ConfigureSCCP Endpoints 35
Task 4: ConfigureSupport for Cisco IP Communicator{Optional) 36
Lab4-1: Implemenling Digit Manipulation 37
Visual Objective 37
Command List 38
Job Aids 33
Dial Plan 38
Task 1: Fix Outbound International PSTN Calling from BR 39
Task 2: Change Internal Numbering Plan to Four-Digit Scheme 40
Task 3: Manipulate Calling Numberin Outbound PSTNCalls 41
Task 4: Manipulate Calling Number and Called Number in Inbound PSTN Calls 42
Task 5: Manipulate Calling Number and Called Number in Intersite VoIP Calls 43
Lab 4-2: Implementing Path Selection 45
Visual Objective 45
Command List 43
Job Aids 45
Dial Plan 45
Task 1: Configure Backup PSTN Path for HQto BRCalling 47
Task 2: Configure Backup PSTNPath for BRto HQCalling 48
Task 3: Configure TEHO at HQ for Calls to North America 49
Task 4: Configure TEHOat BRfor Calls to Europe 49
Lab 4-3 Implementing Calling Privileges 51
Visual Objective 51
Command List 52
Job Aids 52
Task 1. Configure Call Permissions for the HQ Site 53
Task 2: Configure Call Permissions for the BR Site 54
Lab 5-1: Implementing Gatekeepers 55
Visual Objective 55
Command List 55
Job Aids 57
Gatekeeper and Gateway Addressing 57
Task 1 Configure Local Zones and Zone Prefixes 57
Task 2: Configure Gateways to Register withthe Gatekeeper 58
Task 3: Configure Call Admission Control 59
Lab 5-2. Implementing Cisco Unified Border Element 60
Visual Objective 60
Command List 61
Job Aids 61
Task 1: Configure Cisco Unified Border Element Functions 61
Task 2: Configure Codec Transparency 62
Task 3: Configure SIP-to-H.323 Interworking and Media Flows 63
Lab 6-1: Implementing QoS Using Cisco AutoQoS and Manual Configuration 64
Visual Objective 64
Command List 65
Job Aids 65
Task 1: Evaluate VoIP Quality Without QoS Applied 65
Task 2' Configure AutoQoS VoIP 66
Task 3: Fine-Tune QoS Policy 66
Answer Key
Lab 1-1 Answer Key:
Lab 1-2 Answer Key:
Lab 2-1 Answer Key:
Lab 3-1 Answer Key:
Endpoints
Lab 4-1 Answer Key
Lab 4-2 Answer Key
Lab 4-3 Answer Key
Lab 5-1 Answer Key
Lab 5-2 Answer Key
Lab 6-1 Answer Key
Configuring Voice Ports
Configuring DSPs 70
ConfiguringVoIP Call Legs 71
Configuring CiscoUnified Communications Manager Express to Support
73
: Implementing Digit Manipulation 75
: Implementing Path Selection 78
: Implementing Calling Privileges 80
: Implementing Gatekeepers 82
: Implementing Cisco Unified Border Element 83
: Implementing QoS Using CiscoAutoQoS and Manual Configuration 85
68
nqCisco Voice Communications andQoS(CVOICE) v8.0
iv Implementing
CVOICE
Lab Guide
Overview
This guide presents the instructions and other information concerning the lab activities for this
course. You can find the solutions in the lab activity Answer Key.
Outline
This guide includes these activities:
I'ullout Pod Layout and IP Addressing
Descriptionof Simulated PSTN
fab 1-1: Configuring Voice Ports
Lab 1-2: Configuring DSPs
Lab 2-1: Configuring VoIP Call Legs
fab 3-1: Configuring Cisco Unified Communications Manager Express toSupport
Endpoints
Lab4-1: Implementing Digit Manipulation
Lab4-2: Implementing PathSelection
Lab 4-3: ImplementingCalling Privileges
Lab 5-1: Implementing Gatekeepers
Lab 5-2: Implementing Cisco Unified Border Element
Lab 6-1: Implementing QoS Using Cisco AutoQoS and Manual Configuration
Answer Key
Pullout Pod Layout and IP Addressing
Pod Layout and IP Addressing
Podp
1 DHCP
___J)
2 Implementing Crsco Voice Communications and QoS (CVOICE) v8.i 2010 Cisco Systems, Inc
Description of Simulated PSTN
This section describes the functionality ofthesimulated PSTN.
fhe lab lavout consists of two student pods: pod 1and pod 2. Each pod consists of an HQ-x site
and aBR-'x site. These sites are virtually placed in the following ditTerent geograph.cal regions:
TheHQ-x site islocated ina virtual European country.
The BR-x siteis located ina virtual North American country.
This placement determines the dialing rules that are to be followed when dialing to the
simulated PSTN. The HQ-x site follows European dialing rules. The BR-x site follows North
American dialing rules (thatis, theNANP).
Note There is no common European numbering plan in effect within the European Union at the
moment. Therefore, for simplicity, the simulated PSTN supports dialing rules like the NANP.
This will be explained in more detail in a later section. ^__
Student pods are connected to the simulated PSTN using ISDN PRI interfaces. ISDN PRI
interfaces use one oftwo switch types toward thesimulated PSTN.
HQ-1 and 1IQ-2 use switch-type primary-net5.
BR-1 and BR-2 use switch-type primary-ni.
Be aware that using switch-type primary-ni automatically modifies the type of number (TON)
for outbound ISDN calls ifthey have one ofthe following called-party number formats:
Aseven-digit called-party number gels TON subscriber.
A10-digit called-party number gets TON national.
A12-digit called-party number (includes 2-digit country code) prefixed with Oil gets TON
international.
This figure shows DID called-party number ranges that are used to dial into individual sites.
These calls can be placed either from asimulated PSTN phone to asite or from one site to
another via the simulated PSTN. Pods use overlapping directory numbers atthe HQ-x and BR-
x sites.
Lab Guide
1he indi\ idual sites use the DID number ranges that are listed inthis table. Area codes 511.
512. 521. and 522 are assigned to individual sites. Country codes 55 and 66 represent the
virtual European (EL1) and the virtual North American (NA) countries.
Site DID Number Ranges
HQ-mSite (EU)
BR-m Site (NA)
Site internal directory numbers 2XXX
3XXX
Local PSTN DIDrange
555-2XXX
555-3XXX
National PSTN DID range
51m-555-2XXX
52m-555-3XXX
International PSTN DID range
55-51 m-555-2XXX
66-52m-555-3XXX
Note m is your pod number.
Simulated PSTN Phone
Lach ofthe student pods has been using its own simulated PSTN phone: PS'I'N phone I {for
pod 1) and PS'I'N phone 2 (for pod 2).
The simulated PSTN phone is used toinitiate simulated PS'I'N calls tosites ofthe pod towhich
the simulated PSTN phone has been associated, or the simulated PSTN phone is used to
terminate simulated PSTN calls that originated from astudent pod. For this purpose, the
simulated PST Nphone has been using six lines that represent different types ofcalls.
This table outlines which line buttons are available at the simulated PSTN phone as well as
their functions.
ImplementingCisco VoiceCommunications and QoS (CVOICE) v8.0
Lines at the Simulated PSTN Phone
Button Position Button Name*
Function
1
Local
Terminates valid local PSTN calls
2
National
Terminates valid national PSTN calls
3
Interntl
Terminates valid international PSTN calls
4 800
Terminates valid toll-free PSTN calls
5
Premium
Terminates valid premium (900) PSTN calls
6
Emergency
Terminates emergency PSTN calls
names as they appear on the simulated PSTN phone display (button labels)
Thedefinition of valid calls foreach type of call isexplained inlater sections. Inthese later
sections, itwill also be explained how the simulated PSTN phone can be used to initiate calls of
various types.
Caution Placing calls between simulated PSTN phones oftwo pods isnot supported.
Calls Inbound to Simulated PSTN
This section describes which calls arerecognized bysimulated PSTN intheinbound direction
when dialed from student pods andhow thesecallsarerouted.
Calls Inbound from HQ Sites
HQ sites are virtually located in Europe and, therefore, all calls from IIQ sites into the
simulated PS'I'N follow European dialing rules. Forsimplicity, however, variable-length
numbers are implemented for international calls only and not for local and national calls, as is
common in many European countries.
Acall from an HQ site, depending onits called number, iseither terminated at the simulated
PSTN phone orrouted to another site (HQ orBR) ofpod 1orpod 2. The call can also be
returned (hairpinned) lo thesitethat it camefrom.
Thistable lists all the types of valid inbound calls tothe simulated PSTN from HQ sites.
Valid Inbound PSTN Calls from HQ Sites
Valid Called Number Valid TON Call Routed To Calling Number
Presentation Format
112 unknown Emergency line" preserved
[2-9]XX-XXXX (7 digits) unknown Local line* preserved
[2-9]XX-XXXX(7 digits)
subscriber Local line* preserved
0T2-9]XX-[2-9]XX-XXXX(11 digits) unknown National line* preserved
[2-9]XX-[2-9]XX-XXXX (10 digits) national National line* preserved
00-any number of digits unknown Intemtf line* preserved
Any number of digits international Interntl line* preserved
0-800-[2-9]XX-XXXX (11 digits) unknown 800 line* preserved
800-[2-9]XX-XXXX (11 digits) national 800 line- preserved
i 2010 Cisco Systems. Inc
Lab Guide
Valid Called Number Valid TON Call Routed To Calling Number
Presentation Format
0-900-[2-9]XX-XXXX (11 digits)
unknown Premium line* preserved
900-[2-9]XX-XXXX (11 digits) national Premium line* preserved
555-2XXX (7 digits)
unknown HQ-m subscriber
555-2XXX (7 digits)
subscriber HQ-m
subscriber
0-511-555-2XXX(11 digits) unknown HQ-1
national
511-555-2XXX (10 digits) national HQ-1 national
00-55-511-555-2XXX (14 digits) unknown HQ-1 national
55-511-555-2XXX (12 digits) international HQ-1 national
0-512-555-2XXX (11 digits) unknown HQ-2 national
512-555-2XXX (10 digits) national HQ-2 national
00-55-512-555-2XXX (14 digits) unknown HQ-2 national
55-512-555-2XXX (12 digits) international HQ-2 national
00-66-521-555-3XXX (14 digits) unknown BR-1 international
66-521-555-3XXX (12 digits) international BR-1 international
00-66-522-555-3XXX (14 digits) unknown BR-2 international
66-522-555-3XXX (12 digits) international BR-2 international
*I.ines at thesimulated PS'I'N phone thai areassociated with thecall originating pod.
Note Where m is the pod number that originated the call.
Note All calls being routed to pod sites present the called-party number inthe national format,
regardless of the call type.
Calls Inbound from BR Sites
BR sites are \ irtually located in North America. The simulated PSTN follows the NANP
number formats.
A call from a BR site, depending on its called number, is either terminated at the simulated
PSTNphone or routed to another site (HQ or BR) of pod I or pod 2. Ihe call can also be
returned (hairpinned) to the site from which it came.
This table lists all the types of valid inbound calls to the simulated PSTN from BR sites.
Implementing Cisco Voice Communicalions and OoS (CVOICE) v8 0 2010 Cisco Systems, Inc
Valid InboundPSTN Callsfrom BRSites
Valid Called Number
Valid TON
...
Call Routed To
Calling Number
presentation Format
911
unknown
Emergencyline*
preserved
[2-9]XX-XXXX (7digits)
unknown Local line*
preserved
[2-9JXX-XXXX (7digits}
subscriber Local line*
preserved
1-[2-9]XX-[2-9]XX-XXXX(11 digits)
unknown National line*
preserved
[2-9]XX-[2-9]XX-XXXX (10 digits)
national
National line*
preserved
011-any number ot digits
unknown
Interntl line*
preserved
Anynumber of digits
international Interntl line*
preserved
1-800-[2-9]XX-XXXX (11 digits)
unknown
800 line*
preserved
800-[2-9]XX-XXXX (11 digits)
national 800 line*
preserved
1-900-[2-9]XX-XXXX(11 digits)
unknown
Premium line*
preserved
900-[2-9]XX-XXXX (11 digits)
national Premium line*
preserved
555-3XXX (7 digits)
unknown BR-m
subscriber
555-3XXX (7 digits)
subscriber BR-m
subscriber
1-521-555-3XXX (11 digits)
unknown BR-1
national
521-555-3XXX (10 digits)
national BR-1
national
011 -66-521-555-3XXX (15 digits)
unknown BR-1
national
66-521-555-3XXX(12digits)
international BR-1
national
1-522-555-3XXX (11 digits)
unknown BR-2
national
522-555-3XXX (10 digits)
national BR-2
national
011-66-522-555-3XXX (15 digits)
unknown BR-2
national
66-522-555-3XXX (12 digits)
international BR-2
national
011-55-511-555-2XXX (15 digits)
unknown HQ-1
international
55-511-555-2XXX (12 digits)
international HQ-1
international
011-55-512-555-2XXX (15 digits)
unknown HQ-2
international
55-512-555-2XXX (12 digits)
internationa HQ-2
international
at the simulated PSTN phone that are associated with the call originating pod
Where misthe pod number that originated the call
All calls being routed to pod sites present the called-party number in the national format,
regardless ofthe call type.
Lab Guide
Calls Outbound from Simulated PSTN
This section explains hou outbound calls from the simulated PSTN loward student pods can be
placed. r
Predictable Calling-Party Numbers
for placing calls to student pods, each pod is using its own simulated PSfN phone Calling-
party numbers for the simulated PSTN phone-initiated calls are determined depending on
uhich simulated PSTN phone line was selected to originale the call. This figure shows calling-
party numbers and the associated TONs. based on aline selected. Ifacall is originated from the
HOU (toll-free) line, instead offrom the calling number, the calling-party name "PSTN" is
presented (the calling-party number field is erased completely). Ifa call is originated from the
1remium line, the presentation ofthe calling-party number is restricted.
Calling-Party Numbers for Calls
Initiated from Simulated PSTN Phone
PSTN-Phone-x
*?
Local # -
National -
Interntl
800
Premium
Emergency
555444,subscriber
606-555-4444, national
77-606-555-4444, international
"PSTN"
CLIR
112or 911, unknown
To Pod x
Sites
Each line can place any ofthe valid supported call types. The line selection influences the
calling-party presentation only, 'fhe call types and called-party number formats that are valid is
explained in later sections.
Ifno line is selected before dialing at the simulated PSTN phone, and the called number is
dialed immediately, then acall is placed as ifitwas dialed from the Local line.
Valid Outbound Calls from Simulated PSTN Phone
The calls tliat are placed from the simulated PSTN phone are routed to the HQ and BR sites of
pod 1or pod 2. Because HQ and BR sites are virtually located in ditTerent geographical
regions, an outbound call must comply with the dialing rules in that particular region. The
simulated PSTN phone covers both regions as ifit was located in both Europe and North
America at the same time.
This table lists all the call types and their accepted called-party numbers that arc permitted to
dial from the simulated PSTN phone.
Valid Outbound PSTN Calls
Accepted Called Number at Simulated
PSTN Phone
Call Routed To
Called Number
Presentation Format
555-2XXX (7 digits)
HQ-m
subscriber
555-3XXX (7 digits)
BR-m
subscriber
0-511-555-2XXX(11 digits)
HQ-1
national
0-512-555-2XXX(11 digits)
HQ-2
national
1-521-555-3XXX (11 digits) BR-1
national
1-522-555-3XXX(11 digits)
BR-2 national
00-55-511-555-2XXX (14 digits) HQ-1
international
Lab Guide
Accepted Called Number at Simulated
PSTN Phone
00-55-512-555-2XXX (14digits)
00-66-521-555-3XXX (14digits)
00-66-522-555-3XXX (14digits)
011-55-511-555-2XXX(15digits)
011-55-512-555-2XXX (15digits)
011-66-521-555-3XXX (15digits)
011-66-522-555-3XXX (15digits)
Call Routed To
HQ-2
BR-1
BR-2
HQ-1
HQ-2
BR-1
BR-2
Called Number
Presentation Format
international
international
international
international
international
international
International
Note
Where mis the pod number that is associated with the simulated PSTN phone that is
originating the call.
Implementing CiscoVoice Communications and OoS (CVOICE) v8.0
Lab 1-1: Configuring Voice Ports
Complete this lab activity to practice what you learned in the related module.
Activity Objective
In this activity, you will configure voice ports in order to connect a gateway to the PSTN using
digital interfaces. After completing this activity, you will be able to meet these objectives:
Configure a PRI interface for correct signaling, framing, and time slot association
Configure the appropriate ports and dial peers to place calls to the PSTN
Verify the call processing and dial peer matching
Visual Objective
The figure illustrates what you will accomplish in this activity.
Lab 1-1: Configuring Voice Ports
PODP | P- pod number |
Required Resources
Theseare the resources and equipment that are requiredto complete this activity:
A PSTN phone
At least one Cisco Unified IP phone in each site (1 IQand BR)
Lab Guide
Command List
The table describes the commands that are used in this activity.
Cisco 105 Commands
12
Command
ip dhcp pool pool-name
network address mask
default-router IP-address
option 150 ip IP-addres;
telephony-service
ip source-address IP-
address port
max-dn number
max-ephones number
auto assign number to
number
ephone-dn dn-identifier
number telephone-extension
create cnf
controller {tl j el}
slot/subslot/port
forward-digits {num-digi t
] all J extra}
framing {sf | esf
no-crc4}
crc4
linecode {ami | bSzs
hdb3}
clear interface slot/port
clock source {[primary]
line | internal | free-
running}
isdn switch-type {country-
specific- switch -type}
interface {bri | pri}
interface-number
isdn protocol-emulate
{user | network}
Description
Creates DHCP server pool and enters its configuration
mode
Configures network IP address and mask at DHCP server
pool
Configures default gateway IP address at DHCP server
pool
Configures IP address of TFTP server at DHCP server pool
Activates Cisco Unified Communications Manager Express
and enters its configuration mode
Configures source IP address and port for Cisco Unified
Limits maximum number of directory numbers at Cisco
Unified Communications Manager Express
Limits maximum number of IP phones at Cisco Unified
Automatically assigns an already defined directory number
to button 1 of autoregistering Cisco Unified IP phones
Enters IP phone directory number configuration mode
Configures IP phone directory number
Rebuilds configuration of IP phones at Cisco Unified
Enters controller configuration mode
Specifies which digits to forward for voice calls
Selects the frame type for the E1 or T1 controller. T1
options are sf and esf. E1 options are crc4 and no-crc4.
Selects the line code type for the T1 or E1 controller. T1
options are ami and bSzs. E1 options are ami and hdb3.
Resets the specified ISDN interface
Sets the clocking for individual T1 or E1 links
Defines the telephone company switch type
Enters interface configuration mode
Routes all incoming voice calls to the modem and
determines how they will be treated
Defines Layer 2 and Layer 3 user- or network-side
emulation
Implementing Cisco VoiceCommunications and QoS (CVOICE) vS.O
Command
prefix string
pri-group timeslots
timeslot-range [nfas_d
{backup | none ] primary
{nfas_int number j
nfasgroup number | rlm-
group number}} | service]
network-clock-participate
{aim | slot | wic} slot-
number
show controllers el | tl
show isdn status
show voice port
[slot/port;dsO-group
summary]
debug isdn q931
debug voice dialpeer
Description
Specifies the prefix of the dialed digitsfor a dial peer
Specifies an ISDN PRI group on a channelizedT1 or E1
controller
Specifies which clock source to use for DSP clocking
Displays parameters of active voice calls
Displays informationabout the T1/E1 controllers
Displays the status of ISDN interfaces
Displaysconfiguration information about a specificvoice
port or a summary of all voice ports
Examines the ISDN Layer 3 signaling messages
Examines the dial peer operations, including the matching
of inbound and outbound dial peers
Job Aids
Thesejob aids are available to helpyou completethe lab activity.
Lab ISDN PRI requirements for both sites (HQ and BR):
Framing= cyclic redundancy check 4 (CRC4)
Line coding = high-density bipolar 3 (HDB3)
Clock source = line
PRI time slots 1 to 8
Numbering plans:
Internal Numbering Plan
Local HQ Site (EU) Local BR Site (NA)
Internal numbering 555-2XXX 555-3XXX
) 2010 Cisco Systems, Inc. Lab Guide 13
Valid Numbers in Simulated PSTN
Calls from HQ (EU) to PSTN Calls from BR (NA) to PSTN
Local calls NXX-XXXX (7 digits), TOM' unknown NXX-XXXX (7 digits), TON: unknown
NXX-XXXX (7 digits), TON: subscriber NXX-XXXX (7 digits), TON: subscriber
Example. 455-8000 Example: 455-8000
National 0-NXX-NXX-XXXX. TON: unknown 1-NXX-NXX-XXXX, TON: unknown
calls
(0 * 3-digit area + 7 digits) (1 + 3-digit area + 7 digits)
NXX-NXX-XXXX, TON: national NXX-NXX-XXXX, TON: national
(3-digit area + 7 digits) (3-digitarea + 7 digits)
Example: 0-455-455-8000 Example: 1^55-455-8000
International
calls
00 + any number of digits, TON:
unknown
unknown
Any number of digits, TON
international
Any number of digits, TON:
international
Example: 00-23-455-455-8000 Example: 011-23-455-455-8000
Emergency
calls
112, TON: unknown 911, TON: unknown
Note N represents a digit between 2 and 9.
Task 1: Configure DHCP Servers to Support Autoregistration
In this task, vou will configure the DHCP sener al the Cisco Unified Communications
Manager Fxpress thai is miming on your HQ and BR router to support the autoregistration of
\our IP phones.
Activity Procedure
Complete these steps:
Step 1 Configure the DHCP server on your I IQ gateway to support MQ IP phones. Use the
settings from the following table. Fxclusion of'IP addresses has been preconfigured.
14
DHCP Server Settings for IP Phones at HQ
Parameter Value
DHCP pool name HQp-Phones
Network address 10.p.2.0
Network mask 255.255.255.0
Default router 10.p.2.101
Option 150 10.p.250.101
Note p is you pod r umber.
Step2 Configure the DHCPserver on your BRgateway to support BR IPphones. Use the
settings from the following tabic. The exclusion of IP addresses has been
preconfigured.
Implementing Cisco Voice Communications and QoS (CVOICE) vS 0 2010 Cisco Systems, Inc
DHCPServer Settings for IP Phones at BR
Parameter Value
DHCP pool name
BRp-Phones
Network address 10.p.4.0
Network mask 255.255.255.0
Default router 10.p.4.101
Option 150
10.p.250.102
Note p is your pod number.
Step 3 Connect your IP phones to the appropriate ports and observe the DHCP assigned IP
settings.
Note You canverify thesettings at the IP phone by pressing theSettingsbutton andbrowsing
through themenu tocheck theobtained IP parameters. The load file (found in theModel
Information menu) should be basedonSCCP for the phone toauloregister. If ituses a file
that is based on Session Initiation Protocol (SIP), contact the instructor.
Activity Verification
Youhave completed thistask whenyou attainthis result:
YourIPphones received the IPaddress from theDIICPserver.
Task 2: Autoregister Cisco Unified IP Phones
Inthis task, you will autoregister Cisco Unified IP phones atthe Cisco Unified
Communications Manager Express thatis running onyour HQandBRrouter.
Note Thistask involves onlyrudimentary settingsthat enable the verification of PSTN
connectivity. The configuration ofCiscoUnified Communications Manager Expressis
covered in a later exercise.
Activity Procedure
Step1 OnyourHQgateway, enable autoassignment of twodirectory numbers. Configure
the Cisco UnifiedCommunications Manager Express source address using the HQ
loopback 0 interface. Allow five SCCP endpoints and10directory numbers.
Step 2 Onyour \\Qgateway, configure twodirectory numbers with identifiers 1and 2 and
telephone extensions 5552001 and5552002. Thedirectory numbers should support
two parallel calls.
Step 3 Rebuild the configuration files for IP phones at the HQ gateway.
Step4 Onyour BRgateway, enable autoassignment of twodirectory numbers. Configure
the Cisco Unified Communications Manager Express source address using the BR
loopback 0 interface. Allowfive SCCPendpoints and 10directorynumbers.
>2010CiscoSystems, Inc. Lab Guide
Step5 Onyour BRgateway, configure one directory numberwithidentifier 1and
telephone extension 5553001. The directory number should support two parallel
calls.
Step 6 Rebuild theconfiguration files for IPphones at BR galewav.
Step 7 Ensure thai the IPphones successfully register.
You ha\ e completed this task when you attain these resulls:
You verified that the IPphones registered andthatthey obtained a directory number.
Youverified that thetwophones that are registered at 1IQ gateway could call each other.
Task 3: Configure PRI Interfaces
Inthis task, you will configure an ISDN PRItrunk fbreach site (HQand BR)to the PSTN.
Activity Procedure
Step 1 On your HQrouter, ensure that the digital signal processors (DSPs) are clocked
correctly.
Step2 Onyour HQrouter, set the ISDN switch type to primary-net5 (the settingthat is
most common in Europe).
Step 3 Configure an ISDN I'Rl trunk on the El controller. Use time slots I through 8.
Step4 Repeat this procedure to configure the ISDN I'Rl interface on the BRgateway, but
set the ISDN switch type to primary -ni (the setting that is most common in North
America].
You have completed this task when you attain these results:
The ISDN status on Layer 2 at the HQ and BR gateway was
MULTlI'l.EJ-RAMEÊSTABI.ISHED.
The operational state ofthe ISDN I'Rl voice port 0/0/0:15 was dormant (dorm), and the
administrate slate was up.
Task 4: Enable HQ to PSTN Calls
In this task, you will configure dial peers that enable outbound PSTN calls from the HQ site.
The HQ is virtually located in Europe, thus it uses European dialing rules and PSTN access
code 0.
Activity Procedure
Complete these steps on your HQ router;
Step 1 Create three PO I S dial peers that all point to the voice port that corresponds to the
ISDN PRI inlcrfacc. Use these destination patterns:
PSTN access code 0 followed by a seven-digit local number (NXXXXXX.
where N is 2 to 9 and X is any digit). The called number that is sent to the PSTN
must include se\en digits without the PSTN access code. Use dial peer tag 7.
PSTN access code 0 followed by the interarea code 0, the area code (three
digits), and aseven-digit subscriber. The dial peer must match the 12-digit
number (OONXX-XXXXXXX. where Nis 2lo 9and Xis any digit). The called
number that is sent to the PSTN musl exclude the PSTN access code. Use dial
peer tag 10.
PSTN access code 0 followed by an international prefix 00and a variable-length
international number. The dial peer must match any number that slarts with 000.
The called number that issent tothe PSTN must exclude the PSTN access code.
Use dial peer lag 9011.
Step 2 Create two POTS dial peers that point to the voice port corresponding to the ISDN
PRI interface with these destination patterns:
Emergency number 112 {dial peer tag 112)
PSTN access code0 followed by theemergency number 112(dial peertag
1120)
From an HQ phone, call the numbers, using the configured dial peers. The goal is to
make the PSTN phone ring atthe appropriate line. Test the following dialing:
Emergency 112
Local call 4551000 (remember lo addthe PSTN access code0)
National call 4554551000 (remember toadd the PSTN access code 0 and
interarea code 0)
International call 774554551000 (remember toaddthe PSTN access code0 and
international dialing prefix 00). You can avoid waiting for interdigit timeout by
adding #.
Step 4 On your HQ router, debug ISDN Q.931 to inspect the called number that issent to
the PSTN in the previous step. Ifneeded, correct the configuration ofthe dial peers.
Verify that all calls go through.
Step 5 When placing calls, verify the dial peer matching using the debug voice dialpeer
command, and examine active calls using theshowcall active voice command.
Step 6 With ISDN Q.931 debugging active, call an HQ phone from your PSTN phone,
using the number 5552001. You should hear asecond dial tone, and your HQ phone
will not ring. Verify in the debugging output that the call arrived onthe HQ
gateway. In the space that isprovided, write down the PSTN number from which the
call originated (it will be needed in the next task). Why did the call not forward to
theinternal IPphone? Youwill fixtheproblem inthenext task.
Step 3
MM
Youhave completed this taskwhen youattain theseresults:
You successfully placed outbound PSfNcalls from yourHQ phones.
You monitored the call processing using theappropriate debug commands.
Youverifiedthat inbound PSTN calls arriveon the HQgateway but are nol delivered to the
destination (in other words, a phone does not ring).
You examined theparameters of active calls using theshowcall active voice command.
Task 5: Enable PSTN to HQ Calls
Inthis task, you will enable inbound PSTN calls toyour HQ site.
Activity Procedure
Complete these steps onyour HQ gateway-
Step 1 Create aPOTS dial peer for matching inbound PSfN calls, which has the following
characteristics:
It matches calls destined to any number.
It uses direct inward dial.
It has dial peer tag I.
Step 2 From your PST Nphone, call an HQ phone (555200I or5552002), and verify that
the call goes through andrings theappropriate phone.
Step 3 Debug the voice dial peer operations (debug voice dialpeer command), and make
another call from your PS'FN phone toan HQ phone. Which outbound dial peer is
used for the call?
You successfully placed inbound PSTN calls toyour HQ phones.
You monitored the call processing using the debug command, and you found outthat the
\ irtual dial peerthat represents the IPphone directory number wasused.
Task 6: Enable BR to PSTN Calls
In thistask, you will configure dial peers that enableoutbound PS'I'N calls from the BRsite.
The BR site is located in North America, thus it uses the NANP and PSTN access code 9.
Activity Procedure
Complete these steps on your BRgateway-
Step1 Create three POTS dial peers, all pointinglo the voiceport that corresponds to the
ISDNPRI interface. Use these destination patterns:
PSTNaccess code 9 followed by a seven-digit local NANP number
(NXXXXXX. where N is 2 to 9 and X is any digit). The called number that is
sent to the PSTNmust include seven digils without the PSTNaccess code. Use
dial peer tag 7,
PS'I'N access code9 followed by a validNANP long-distance numberthe long
distance prefix L the area code (three digits), and a seven-digit subscriber
{l NXX-NXX-XXXX. where N is 2 to 9 and X is any digit). The called number
that is sent to the PSTN must excludeIhe PSTN accesscode. Usedial peer tag
10.
PSTN access code 9 followed by an international prefix Ol I and a variable
length international number. The called number thai is sent to the PSTN must
exclude the PSfN access code. Use dial peer tag 9011.
Implementing CiscoVoice Communications and QoS (CVOICE) vS.O 2010CiscoSystems, Inc.
Note
Outbound calling from the BR site to the international PSTN number will not work until you
reach the Lab4-1 activity. ^
Step 2 Create two POTS dial peers that point to the voice port corresponding to the ISDN
PRI interface with these destination patterns:
Emergency number 911 (dial peer tag 911)
PSTN access code 9 followed by the emergency number 911 (dial peer tag
9911)
Step 3 From the BR phone, place test calls to the numbers that are configured in the dial
peers. The goal is to make the PSTN phone ring at the appropriate line. Remember
to dial with the correct PSTNprefix.
Step 4 Use ISDN Q.931 debugging to inspect the called number that is sent to the PSTN. If
needed, correct the configuration ofdial peers. Verify that all calls go through.
Step 5 When placing calls, verify the dial peer matching using the debug voice dialpeer
command, and examine active calls using the show call active voice command.
You successfully placed outbound PSTN calls from your BR phone.
You monitored the call processing using the appropriate debug commands.
You examined the parameters ofactive calls using the show call active voice command.
Task 7: Enable PSTN-to-BR Calls
In this task, you will enable inbound PSTN calls to your BR site.
Activity Procedure
Complete these stepsonyourBRgateway:
Step 1 Create aPOTS dial peer for matching inbound PSTN calls, which has the following
characteristics:
It matches calls destined to any number.
Il uses direct inward dial.
It has dial peer tag 1.
Step 2 From your PSTN phone, call the BR phone (5553001), and verity that the call
succeeds.
Step 3 Verity' call routing on the BR gateway using the appropriate debug commands.
Youhave completed thistaskwhen youattain these results:
You successfully placed inbound PSTN calls toyour BRphone.
You monitored the call processing using theappropriate debug commands.
Lab Guide
Complete this lab activity topractice what you learned in the related module.
Activity Objective
In this activ ity. you will configure DSPs for voice termination and verily their operations. After
completing this activity, you will be able tomeet these objectives:
Configure DSP resources to terminate voice calls
Configure DSPs to support appropriate codec complexitv
Verify the use of DSP resources
Visual Objective
The figure illustrates what you will accomplish inthis activitv.
PODP
HO Phonal
555-2001
HQ Phone 2
S55-2002
Required Resources
PSTN Phone-p
(Cisco UnifiedIP phone)
P = pod number
These are the resources and equipment that arc required to complete this activity:
A PSTN phone
Two Cisco Unified IP phones in your HQsite
20 Implementing Cisco VoiceCommunications and QoS (CVOICE! v8 0
Command List
Cisco IOS Commands
Command
voice-card voice interface
slot
Description
Enters voice card configuration mode
codec complexity flex ]
high | medium | secure
Specifies the maximum codec complexity that is supported
by thevoice card DSPs ^
codec codec-name
show voice dsp
Specifies the codec that isused on the IP phone
Displays the DSP resource usage and call parameters
show voice call status
Provides a compact view ofactive call legs and their
parameters ___
Provides detailed information about active voice calls
Job Aids
These job aids are available to help you complete the lab activity.
Local HQ Site (EU)
Local BR Site (NA)
Internal numbering
555-2XXX
555-3XXX
Calls from HQ(EU)to PSTN
Calls from BR (NA) to PSTN
Local calls
NXX-XXXX (7 digits), TON: unknown
NXX-XXXX (7 digits), TON: subscriber
Example: 455-8000
NXX-XXXX (7 digits}, TON: unknown
Example: 455-8000
National
calls
0-NXX-NXX-XXXX, TON: unknown
(0 + 3-digit area + 7 digits)
NXX-NXX-XXXX. TON: national
(3-digit area + 7 digits)
Example: 0-455-455-8000
1-NXX-NXX-XXXX,TON: unknown
NXX-NXX-XXXX, TON: national
(3-digit area + 7 digits}
Example. 1-455-455-8000
International
calls
00 +any number of digits, TON:
unknown
Any number of digits,TON:
international
Example: 00-23^55-455-8000
unknown
Anynumber of digits, TON:
international
Example: 011-23-455-455-8000
Emergency
calls
112, TON: unknown
911, TON: unknown
Note
Nrepresents a digit between 2 and 9.
)2010Cisco Systems, Inc.
Lab Guide
1hese are gateuay codec complexities:
l-0u:G,7ll.clearchannel
Medium: G.729A. G.729AB. G.726. G.722, and Fax Relay
High: G.723.1. G.728. G.729. G.729IJ. iLBC. Modem Relay
Task 1: Operate DSPs in Default Codec Complexity Mode Flex
In this task, you will test DSP operations in default codec complexity mode flex.
Activity Procedure
Complete ihese steps onyour HQ gateway-
Step 1 View the automatically created ephones using the show ephone command. There
should be two ephones with extension numbers that are associated with the ephones.
View the preferred codec of each ephone. It should be"7l lulaw."
Note
The ephones are created as aresult ofthe autoregistration when the endpoints first attempt
toregister. Autoregistration is enabled bydefault on Cisco Unified Communications
Manager Express
Step 2 from an [IQ phone, call a PS'I'N local number (0-NXXXXXX, where Nis 2to9
and Xisany digit), and answer the eall onthe PSTN phone.
Step 3 Verify the codec that is used on the VoIP call leg between the HQ phone and the HQ
gateway, using two methods:
I[se appropriate show commands onthe iIQ gateway.
On the IPphone, press theSettings button and choose Status>Call Statistics.
Note Both methods should show that the VoIP call leg uses G.711 with a 20-ms payload size.
Step 4 On the IIQ router, change the preferred codec ofthe first ephone (5552001) to
il.BC. and reset the phone usingthese commands:
ephone 1
reset
Step 5 From the first IIQ phone, call the PSTN phone (0-NXXXXXX). and check the
codec that is used during the call. Display DSP usage using different options ofthe
showvoice dsp command, including thedetailed keyword.
Step 6 fromthe PSTN phone, call the first HQ phone (555-2001}. andcheck thecodec that
is used inthe call. It shouldbethe sameas the codecthat is used inthe call that
originated in HQ.
Step 7 On the HQ router, enter the voice-card 0conllguration mode, and attempt tochange
the DSP codec complexity to medium. This operation should fait, due tothe existing
\oice ports.
22 Implementing Cisco Voice Communications and QoS (CVOICE) v8.0 2010 Cisco Systems Inc
You verified thatthe flex mode supported high-complexity codecs such as iLBC.
You viewed the DSP usage for voice termination using different options ofthe show voice
dsp command.
Task 2: Operate DSPs in Medium Codec Complexity Mode
In this task, you will configure the voice card to support low- and medium-complexity codecs,
and you will verify the DSP operations.
Activity Procedure
Complete thesesteps on your HQrouter:
Step 1 Perform this procedure todelete the voice port:
Enter the voice port configuration mode (for example, voice-port 0/0/0:15). and
shut down the port.
Enter the controller configuration mode (forexample, the controller el 0/0/0),
and shut down the controller.
Inthecontroller configuration mode, remove the PRI group usingthe no pri-
group command.
Note This will also remove the port command fromall POTSdial peers.
Step 2
Enter thevoice cardconfiguration, andchange thecodec complexity to medium.
Step 3 Perform this procedure to re-createthe voiceport:
Enter thecontroller configuration mode, anddefine the ISDN PRI group using
the pri-group timeslots 1-8 command.
Activate the controller using the no shutdown command.
Reapply thevoice porttoalldial peers (forexample, add port 0/0/0:15
command to the dial peers).
Step4 From thefirst HQphone (preferring iLBC), call a PSTN local number (0-
NXXXXXX), and checkthe codecthat is used inthe call. The codecshouldbe
downgraded toa lower-bandwidth, medium-complexity codec. Write down which
codec is used, in the space that is provided:
____ Step 5 With one active conversation, call a PSTN national number (0-0-NXX-XXXXXXX)
from the secondHQphone (preferring G.711). Checkthe codecthat is usedin the
call. Display DSP usage, usingvarious options ofthe showvoicedsp command.
Step6 Change thecodec complexity onthe HQ gateway backto flex mode. Remember to
reapply the voice port to al! POTS dial peers.
2010CiscoSystems. Inc. Lab Guide 23
Youhave completed this task when you attainthese results:
You verified that the iLBC codec was nolonger supported forvoice tennination when the
DSP resources were configured for medium codec complexity.
You \erilled that the codec negotiation for voice termination resulted in a codec that did
not exceed the permitted complexity and that it did not consume more bandwidththan the
preferred one.
Implementing Cisco Voice Communications and QoS (CVOICE) v8 0 201C Cisco Systems, Inc
Activity Objective
In this activitv. vou will configure VoIP dial peers and tune gateway parameters to allow VoIP
calls between your HQ and BR sites. After completing this activity, you will be able to meet
these objectives:
Configure basicVoIPdial peer parameters
Configure codecs and codec classes, and verify codec negotiation
Configure and verify H.323 fast start and slow start
Configure theH.323 gateway interface bind feature
Visual Objective
The figure illustrates what you will accomplish inthis activity.
HQPimnm
555-2001
^
PODP
HQPhone2
SS5-2002
Required Resources
| P=pod number |
These are theresources andequipment that arerequired to complete this activity:
"Iwo Cisco Unified IP phones in your HQ site
One Cisco Unified IP phone in your BR site
Lab Guide
Command List
Ihc table deseribes the commands that are used in this activity.
Cisco IOS Commands
Command
codec codec-name
codec preference 1-14
codec-name
dial-peer voice tag voip
h323-gateway voip bind
sreaddr
session target
ipv4:x.x.x.x
voice class codec tag:
voice service voip
h323
sip
Description
Specifies which codecis tobe usedfor callsmatching this
dial peer
Configures one entry inthe codec list under the voice
class codec command. Repeatthis command as many
times as you need, to specify codecs in this list.
Enters dial-peer configuration mode and specifies VoIP
Configures the H.323 gateway interface bind feature
Configures the VoIP dial peer to use SIPsignaling
Specifies the destination IPv4 address forthe gateway
terminating a VoIP call
Entersvoice class codec configuration mode. In dial peer
mode, it attaches the voice class to the dial peer.
Enters H.323 mode from voice service voip configuration
mode
Enters SIP mode from voice service voipconfiguration
mode
call start slow | fast
Defines H.323 slow start or fast start in H.323 mode
bind control source-
interface
Configuresthe interface bindfeature for SIP signaling in
SIP mode
Displays information on active calls
(tag)|(summary)
Displays dial-peer configuration information
show dialplan number
number
Displays which dial peers are matched when a particular
telephone number is dialed
Displays the status of active voice calls
debug voice dialpeer
Monitors the dial peer matching process
debug voip ccapi inout
Displays real-time call control processing and call leg
information
debug h22 5 event Monitors H.225 events
debug h245 event Monitors H.245 events
debug h245 asnl Monitors H.245 ASN.1 library
debug ccsip message Displays SIP signaling messages
Job Aids
Local HQ Site (EU)
Local BR Site (NA)
Internal numbering
555-2XXX
555-3XXX
Task 1: Configure Basic VoIP on the HQ Gateway
In this task, you will configure abasic VoIP dial peer for HQ-to-BR connectivity.
Activity Procedure
Complete this steponyourHQgateway:
Step 1 Create a VoIP dial peer with these parameters:
It matches the entire 5553XXX range.
It points to loopback 0ofyour BR gateway (10.p.250.102, where pis your pod
number).
It disablesvoice activity detection(VAD).
It has dial peer tag 3000.
On your BR gateway, enable dial peer debugging. Make sure that your Telnet connection
has configured the terminal monitor.
From your first HQ phone, call the BR phone (555300I) and answer the call. Examine the
debugging output, and confirm that no inbound dial peer has been matched on the BR
gateway.
With the call still active, view the active calls using the show voice call status command to
verify thatthe inbound dial peeris 0.
With the call still active, display its VAD setting using the show call active voice | inc
VAD command. Compare the VAD setting on both gateways. The setting should be
different.
With the call still active, use the appropriate show command toexamine the codec that is
used in the call. End the call.
Call the BR phone (5553001) from your second HQ phone. Answer the call and examine
thecodec. Youshould seethatbothcallsuse G.729, despite the fact thatthepreferred
codec ofthe first HQ phone isiLBC, and the preferred codec ofthesecond HQ phone is
G.711.
View thedefault codec ofthe VoIP dial peer,using theappropriate commands. Youmay-
want toselectively display the parameter using theshow dial-peer voice 3000 | inc codec
command.
Lab Guide
Task 2: Configure Codec onthe HQ Gateway
In this task, you will configure asingle codec in the VoIP dial peer on your HQ gateway.
Activity Procedure
Complete these steps onyour HQ gateway:
Step 1 Setthe VoIP dial peer (3000) codec tog723r53.
Step 2 Imm an HQ phone, call yourBR phone (555300I) and answer the eall. The call
should disconnect as soon as it is answered. This behavior is typical for acodec
mismatch.
Step 3 I se H.245 debugging to \ iew the codec negotiation between the gateways. Has Ihe
codec negotiation between thegateways been successful?
Note The Cisco Unified IP Phone 7965 supports these major codecs: G.722, G.711, G.729, and
iLBC.
Youhavecompleted this task when you attainthese results:
You verified that the H.245 codec negotiation was successful, 'fhe debug h245 asnl
command displayed the codec proposal that was sent by the HQ gateway (g7231). which
was accepted by the BRgatewav.
You identified that the reason for the call disconnect musl be thai theCisco Unified IP
phones do not support ihe G.723 codec.
Task 3: Configure Asymmetric Codec Negotiation
In this task. >ou will configure multiple codec proposals and examine their negotiation.
Activity Procedure
Complete thesestepson your HQ and BRgateways:
Step 1 OnyourHQ gateway, create a codec class with thispreference order:
first codec: g723r53
Second codec: ii.BC
Third codec: g729br8 (Annex B)
Step 2 On\our HQgateway, remove theG.723 codec from theVoIP dial peer(3000). and
attach the codec class to the dial peer.
Step3 Onyour BRgateway, createa codecclass withthis preference order:
First codec: iLBC
Second codec: g723r53
Third codec: g729br8 (Annex B)
Step4 Onyour BRgateway. configure a VolPdial peer (2000) withtheseparameters:
The destination patternshouldmatch the DID rangeof HQphones (5552XXX).
Thesessiontarget shouldpoint toyour IIQgateway loopback 0 address
(I0.p.25().l()l. where p is your pod number).
28 Implementing Cisco Voice Communications and QoS(CVOICE) v8.0 2010CiscoSystems, Inc
Ml
It should have an associated codec class that is defined onyour BR gateway in
tlie previous step.
Step 5 From an HQ phone, call your BR phone (5553001), and answer the call. The call
should work. Examine the inbound dial peer selection on your BR gateway. Check
the negotiated codec using the debug h245 asnl command.
Step 6 From your BR phone, call an HQ phone (5552001), and answer the call. The call
*" should disconnect. Use available debug methods to examine the cause ofthe failure.
mm Activity Verification
*" You verified that inbound dial peers were successfully matched based on the destination-
pattern command. Ihe debug voice dialpeer command could be used to validate dial-peer
<mm matching.
You examined codec negotiation using the debug h245 asnl command. You verified that
* the codec list was sent by the originating gateway tothe tenninating gateway. The
terminating gateway looks up its prioritized codec list and chooses the first codec, if
____ available, from the received H.245 proposal. The HQ gateway selects g723r53. which is
not supported by IPphones, and the call fails.
_M You confirmed that the codec selection was based on the priorities that were configured on
the tenninating gateway.
Task 4: Configure Symmetric Codec Negotiation and Examine
. Fast Start
In this task, you will configure multiple codec proposals to make VoIP calls work in both
mm directions.
Activity Procedure
Complete these stepson yourHQandBRgateways:
__ Step 1 On your HQ and BR gateways, modify the codec class to use these codecs only, in
this preference order:
g First codec: g729br8 (Annex B)
Second codec: iLBC
mU Step 2 Make sure that the codec class isattached tothe VoIP dial peers onboth gateways.
Step3 Test VoIP callsin both directions. The callsshould work.
Step 4 Use the debug h245 events command toensure which H.323 call setup method (fast
start or slowstart) is usedby default on Ciscovoice gateways.
mW> You have completed this task when you attain these results:
You verifiedthat VoIPcalls workedbetween the phones in the HQand BRsites. Youmay
mtt have performed some debugging to examine the dial-peer matching and 11.245 codec
negotiation.
Youfound that Ciscogateways used the fast-start methodbydefault.
Lab Guide 29
Task 5: Configure Slow Start and the Interface Bind Feature
In this task, you will change the default fast-start method to slow start and configure the
interlace bind feature.
Activity Procedure
Step 1 Configure your IIQ gateway lo use slow-start signaling. Leave your BR gateway at
the default setting of fast-start signaling.
Step 2 Make calls between the sites. Use appropriale debug commands on both gateways to
ascertain which signalingmethod is usedin whichdirection. Lxamine the slow-start
signaling process.
Step 3 On both gateways (HQ and BR), source the H.323 signaling traflic from the
loopback 0 interfaces.
Step 4 From an HQ phone, call your BR phone. Use the debug h245 events command on
yourBRgateway toverify that thesignaling messages aresourced from the
loopback 0 IP address.
You havecompleted this task when you attainthese results:
You verified that slow start was used for HQ-originatcd calls, and fast start was used for
BR-originated calls.
You examined the slow-start process and confirmed that 11.245 negotiation started allertlie
call was answered and consisted of three main phases: exchange of'TCSs. master and slave
determination, and OLCexchange.
You verified that the 11.323 gateway interface bind feature caused the signaling packets to
be sourced from the defined interface. This can beconlinned with the debug h225 events
command.
Task 6: Configure SIP Signaling
Inthistask, youwill use SIPas the VoIP signaling protocol.
Activity Procedure
Complete these steps on your HQgatewav:
Step 1 Create VoIP dial peer 3001 with the same destination target and destination pattern
as dial peer 3000. Configure the newVoIPdial peer to use SIPv2.
Step 2 Set the preference ofthe VoIP dial peer 3000 to a worse value, to make it the second
choice.
Step3 From an HQ phone, eall your 13R phoneand use the appropriate debugcommand to
examine the SIPmessageexchange. Based on the SDPbody, decidewhich offer
mechanism (early offeror delayed offer) is used bydefault on Cisco gateways, and
write it in the space that is provided:
Step 4 OnyourHQgateway, configure thebind interface feature forSIPsignaling and
media. Source the traffic from the loopback 0 IP address.
Implementing CiscoVoice Communications and QoS (CVOICE) vS.O 2010 Cisco Systems. Inc
You examined the SIP signaling using the debug ccsip messages command.
You verified that Cisco gateways use SIP early offer by default, by identifying that the SDP
body, which contains the codec proposals, was carried in the INVITE message.
You shut down the SIP dial peer on your HQ gateway to make sure that H.323 is used in
the next exercises.
2010 Cisco Systems, Inc. Lab Gulde
Lab 3-1: Configuring Cisco Unified
Communications Manager Express to Support
Endpoints
Complete this lab activitv lopractice what you learned in ihc related module.
Activity Objective
In this activitv. vou will implement SCCP endpoints. The HQ site will host two SCCP phones
and. optionally. one Cisco IP Communicator. The BR site will host one SCCP phone. After
completing this activity, vou will beable to meet these objectives:
Configure SCCP-related Cisco Unified Communications Manager Express parameters
Implement SCCP endpoints
Visual Objective
The figure illustrates what you will accomplish inthis activitv.
Lab 3-1: Configuring Cisco Unified Communications
Manager Express to Support Endpoints
^
HQ Phone 1
555-2001
HQPhons2
555-2002
BR Phone SCCP
555-3001
PODP
PSTN Phono p
(Cisco Unified IP phono)
P = pod number
Required Resources
32
These are the resources and equipment that arc required tocomplete this activity:
Two Cisco I'nified IP phones in your IIQ site
One Cisco Unified IP phone in your BR site
Implementing Cisco VoiceCommunications and OoS (CVOICE) vS.O
& 2010 Cisco Systems, Inc.
Command List
The table describes the commands that arc used inthis activity.
Cisco IOS Commands
Command
ip dhep pool
max-dn
max-ephones
load
time-format
date-format
type
telephony-service
(no) auto-reg-phone
cnf-file location tftp: |
flash:
cnf-file perphone
perphonetype
create cnf-files
ephone ephone-id
mac-address
button button-index;dn-
index
show telephony-service
show ephone
debug tftp event
debug ephone register
Description
Defines a DHCP pool ofaddresses that must include a
network, default router, and option 150
Defines the maximum numberofdirectory numbers that
areconfigured in telephony-service configuration mode
Defines the maximum numberof SCCPendpoints that are
configured in telephony-service configuration mode
Defines the binding oftheimage filename tophone mode
thatisconfigured in telephony-service configuration mode
Defines the time format that is displayed byendpoints that
isconfigured in telephony-service configuration mode
Definesthe date format that is displayed byendpoints that
isconfigured in telephony-service configuration mode
Defines thetype oftheendpoint thai isconfigured in
ephone configuration mode. Needed toidentify the desired
firmware image.
Enters configuration mode forSCCP-based Cisco Unified
Controls autoregistration of SCCP endpoints
Defines the configuration file location forSCCPendpoints
Defines thegranularity ofconfiguration files forSCCP
endpoints
Writes configuration files into the configured or default cnf-
file location
Definesan SCCP endpoint with an identifier and enters
ephone configuration mode
Configures an endpoint MAC addressthat isconfigured in
ephone configuration mode
Associatesa phonebutton with a directory number index.
Multiple separator optionsexist, inaddition to ':'
Displays the SCCP-related Cisco Unified Communications
Manager Expresssystemparameters. Multiple options are
available to examine specific settings.
Displays the status of registered SCCPendpoints
Displays the voice dialpeers. Thesummary option
provides a brief output.
Monitors TFTP server events
Monitors SCCP endpoint registration
Lab Guide 33
Job Aids
Thesejob aids are available to help you complete the lab activity.
Local HQSite (EU)
Local BR Site (NA)
Internal numbering
555-2XXX
555-3XXX
Task 1: Delete Existing SCCP Endpoints on HQ Gateway
In this task, you will remove existing SCCP endpoints and block autoregistration on your HQ
gateway.
Activity Procedure
Complete these steps onyour IIQ and BR gateways:
Step 1 Disable autoregistration and autoassignment ofSCCP endpoints.
Step 2 Delete the autoconfiguredephones.
Step 3 Delete the manually defined ephone-dns.
You have completed thistaskwhen you attain these results:
You displayed the existing ephones to make sure that all auUnregistered ephones had been
deleted.
You verified that all phones reported "registration rejected."
Task 2: Configure Cisco Unified Communications Manager
Express System Parameters for SCCP Endpoints
In this task, you will configure general system parameters for SCCP-related CiscoUnified
Communications Manager Express functionality onthe HQ and BR gateways.
Activity Procedure
Complete thesesteps on your HQgateway:
Step 1 Verify that the DHCP service is running onthe HQ gateway, including option 150.
The Cisco Unified Communications Manager Fxpress should use the loopback 0
address for communications with endpoints.
Step 2 Enable IPv4 and IPv6 operations inthe telephonv service mode, with preference to
IPv4.
Step 3 Configure a European-style daleandtime formal to bedisplayed on theSCCP
endpoints (dd-mm-yy and 24h).
Step 4 Set the location of endpoint files to Hash and allow per-phone configuration files.
Step5 Perform the sameprocedure on your BRgateway to configure Cisco Unified IP
phone systemsettings, with this exception:
Configure the North American dateand time formal to be displayed on the
SCCP endpoints {mm-dd-yy and I2h).
Implementing Cisco Voice Communications and QoS(CVOICE) v8.0 2010CiscoSystems, Inc.
You have completed this task whenyou attainthis result:
Ifyou display the systemwide telephony-service settings, you can verify that the desired
parameters are in place.
Task 3: Configure SCCP Endpoints
In this task, you will configure one SCCPendpoint ineach site.
Activity Procedure
Step 1 OnyourHQ gateway, create five dual-line SCCP directory numbers with extensions
5552001, 5552002, 5552011. 5552012, and 5552003.
Step 2 Onyour HQ gateway, configure anephone [I] with these parameters:
Usethe MAC address of your first HQphone.
Attach ephone-dn with extension 5552001 to the first button.
Attach ephone-dn with extension 5552011 to the second button.
Select theappropriate phone type(7965).
Step 3 On your HQ gateway, configure anephone [2] with these parameters:
Use theMAC address of yoursecond HQphone.
Attach ephone-dn with extension 5552002 to the first button.
Attach ephone-dn with extension 5552012 to the second button.
Attach ephone-dn with extension 5552003 to the third button.
Select the appropriate phonetype (7965).
Step 4 On your HQ gateway, enable debugging ofthe ephone registration process.
Step 5 Onyour HQ gateway, re-create theconfiguration files.
Note Phone reset can be triggered either by using the reset command in ephone configuration
mode, or by pressing the Settings button and entering the sequence **#** on the phone
keypad.
Step 6 Repeat the procedure to configure an SCCP endpoint on your BR gateway, with the
extension number 5553001.
Youhave completed thistaskwhen youattain theseresults:
You examined the booting process ofthe SCCP endpoints.
You verified successful phone registration atthe respective Cisco Unified Communications
Manager Express devices.
You verified that the SCCP phones could call each other within the HQ site and through the
IP WAN.
' 2010 Cisco Systems, Inc
Lab Guide 35
Task 4: Configure Support for Cisco IP Communicator
(Optional)
Inthistask, vou will configure the llQ-based Cisco Unified Communications Manager Express
to support the Cisco IP Communicator.
Activity Procedure
Complete these steps on your HQgateway:
Stepl Install Cisco IP Communicator onyour local computer. Your computer must have IP
connectiv ity tothe loopback 0 address of the IIQ gateway. Depending onyour lab
setup. >oumay have to update therouting onyour computer or plug your computer
into an HQ phone.
Step 2 Start Cisco IP Communicator and select Menu >Preferences >Network.
Step 3 In the Dev ice Name section, choose the desired network adapter, and note the device
name in the fonnat SEP<adapter-mac-address>.
Step 4 In the 11 TP Servers section, select Use these TFTP servers:, and enter the IIQ
gatew ay loopback 0address (10.p.250.101. where pisyour pod).
Step 5 On vour HQ gateway. create anew ephone-dn with the extension 555-2004 and a
new ephone wilh the MAC address that was noted in Step 3and the extension that is
attached to its first button.
Youhave completed thistask when you attain these results:
You verified that Cisco IP Communicator registered with Cisco Unilied Communications
Manager Express.
Youverified that callsworked between Cisco IPCommunicator andotherphones.
36
Lab 4-1: Implementing Digit Manipulation
Activity Objective
In this activity, you will change the internal numbering plan lo a four-digit scheme and
configure digitmanipulation to allow PSTN andintersite calls. After completing this activity,
you will be able to meet these objectives:
Configure digit manipulation for inbound PSTN calls
Configure digit manipulation for outbound PSTN calls
Configure digit manipulation to enable intersite VoIP calls, using an intersite prefix andsite
codes
Visual Objective
Thefigure illustrates whatyouwill accomplish in thisactivity.
Lab 4-1: Implementing Digit Manipulation
PODP
' P- pod number |
Required Resources
These are the resources and equipment that are required to complete this activity:
A PSfN phone
Two CiscoUnified IP phones inyour HQsite
One CiscoUnifiedIPphonein your BRsite
Lab Guide 37
Command List
The table describes the commands thai are used in this activity.
Digit Manipulation Commands
Command
Description
digit-strip
Strips all the digits that explicitly match the POTS
dial peer. Digit stripping is enabled by default on
POTS dial peers.
prefix digits
Specifies the prefixof the dialed digits for a dial peer
forward-digits [0-32]| all | extra
Specifies which digits to forwardfor voice calls
num-exp dialed-digite substitution
Defines how to expand a telephone extension
number into a particular destination pattern
voice translation-rule rule tag
Defines a voice translation rule for voice calls
rule precedence /match/ /replace/
[type {match-type replace-type J
[plan {match-plan replace-plan)jJ
Defines a rule within a voice translation rule
voice translation-profile profile-
name
Specifies a translationprofile forall incoming VoIP
calls
translate {called | calling |
redirect-called} translation-rule-
number
Associates a translation rule with a voice translation
profile
translation-profile {incoming |
outgoing} name
Assigns a translation profile to a dial peer
test voice translation-rule number
input-teet-string [type match-type
[plan match-type]
Tesls the functionality of a translation rule
debug isdn g931
Monitors ISDN Q931 signaling
debug voice translation
Monitors translation operations
Job Aids
These jobaids are av ailable tohelp you complete the lab activity.
Dial Plan
The table represents the dial plan that will beused inthelabs.
Site Internal Numbering Plan and PSTN DID Ranges
HQ Site (EU)
BR Site (NA)
Internal extensions
2XXX
3XXX
Site codes 810
820
PSTN access code 0
9
Local DID range
555-2XXX
555-3XXX
National DID range
51p-555-2XXX
52p-555-3XXX
International DID range
55-51 p-555-2XXX
66-52p-555-3XXX
Implementing Cisco Voice Communications and QoS (CVOICE] v8 0
mm
Note
p: 1 to 2 (pod number)
Calls from HQ(EU) to PSTN
Local calls
Example: 455-8000
Example: 455-8000
National
calls
(3-digitarea + 7 digits)
Example: 0-455-455-8000
Example: 1^55-455-8000
International
calls
unknown
international
Example; 00-23-455-455-8000
unknown
international
Example: 011-23-455-455-8000
Emergency
calls
112, TON: unknown
911, TON: unknown
Note
N represents a digit between 2 and 9.
Task 1: Fix Outbound International PSTN Calling from BR
In this task you will fix tlie issue ofISDN switch-type primary-ni, which isused at the BR
gateway. The primary-ni automatically modifies the called-party number for outbound calls, if
it is inNANP format (011 followed by 12 digits, which includes a two-digit country code). The
primary-ni changes the type ofnumber (TON) to international. Such acalled-party number is
not compliant withthesimulated PSTN that is used in thelab.
Note
Ifthe 011 prefixis used, the correct TONshould be unknown.
Activity Procedure
Complete these steps on your BR gateway:
Stepl Enable ISDN Q931 debugging at the BRgateway. Make surethat theterminal
monitor is configured.
Step2 Place a PSTN international call fromyour BRphone. For example, dial 9-011-77-
455-455-1000#, where 9 is the PSTNaccess code and # prevents waiting for the
interdigit timeout. This call will fail. Observethe type of numberthat was
automatically set for this call by the ISDNprocess at the BRgateway.
Step3 Createa translation rule (usingtag I) and translation profilethat set the type of
number to unknown fbr all called-party numbers that start with 90! I.
Step 4 Associate the translation profile with the ISDN PRI voice port in the outbound
direction.
Step 5 Place the same test call again, and it should succeed.
Lab Guide 39
Youhave completed thistask when you attaintheseresults:
You observed lhai. if youaredialing international PSTN number in NANP format, the
primary-ni automatically sets TON international.
You fixed thisbehavior ofthe primary-ni by using a voice translation ruleandvoice
translation profile.
Task 2: Change Internal Numbering Plan to Four-Digit Scheme
In this task, vou will convert the internal numbering plan lofour-digit schemes: 2XXX inthe
HQ site and 3X.XX in the BR site.
Activity Procedure
Complete these steps onyour [IQ andBRgateways:
Step 1 On your [\Qgatewav. change the extension number of your SCCP endpoints. using
this procedure:
Modify the ephone-dns:
First HQ phone: 555-2001 to 2001 and 555-2011 to 2011
Second HO phone: 555-2002 to 2002. 555-2012 to 2010, and 555-2003
to 2003
Optional HQ Cisco IP Communicator: 555-2004 to 2004
Restart the endpoint (in ephone mode).
Step 2 Onyour BR gateway, change theextension number of your SCCP endpoint. using
ihis procedure:
Modifv the ephone-dn
BR phone: 555-3001 to 3001
Restart the endpoint (in ephone mode).
You verifiedthat the endpoints had reregisteredand obtained four-digitextension numbers
in the respective range.
You placed test calls and verified that intrasite calls continued to work using the modified
numbers. You verified that the intersite calls no longer worked, due to the updated
numbering scheme.
Fromyour HOand BRphones, you called the PSTNnumbers and viewed the calling
number that was displayed on the PSTN phone. Ihc calling number was a four-digit
number that does not allow callback.
Task 3: Manipulate Calling Number in Outbound PSTN Calls
In this task, you will configure digit manipulation to properly present the calling number when
placing callstothe PSTN.
Activity Procedure
Step 1 Manipulate the calling number in outbound PSTN calls on your HQ gateway:
Configure the appropriate translation profiles and rules to convert the calling
number from 2XXX to 5552XXX.
Apply the translation profile to the ISDN PRI voice port in the outgoing
direction. Alternatively, you could apply the translation profile tothe
appropriate POTS dial peers.
Step 2 Manipulate the calling number in outbound PSTN calls on your BR gateway:
Configure the appropriate translation profiles and rules to convert the calling
number from 3XXX to 5553XXX.
Note You will need to reuse the voice translation profile that was configured in Task 1, asyou can
associate only one translation profile with asingle voice port in one direction!
Make sure that the translation profile isapplied tothe ISDN PRI voice port in
the outgoing direction.
Step 3 Debug the translations using the appropriate Cisco IOS commands.
Note Simulated PSTN works differently from real PSTN. When calling national or international
PSTN numbers in the classroom, the PSTN phone displays the calling number as a local
number that isconfigured in this task. Real PSTN would prefix the correct area and country
codes, depending onhow farthecall wassent.
You tested the translation rule by using the appropriate Cisco IOS commands.
You placed PSTN calls from your HQ and BR phones and verified that the calling number
was presented using thecorrect DID siterange.
You enabled appropriate debugging on your gateways and monitored the translation
operations.
Note If you want to adjust the calling numbers for the simulated PSTN, you can optionally
configure multiple translation profiles and apply them to the appropriate outbound dial peers
(local, national, and iniernational). The correct area or country code can Ihen be prefixed,
based on the selected destination.
Lab Guide
Task 4: Manipulate Calling Number and Called Number in
Inbound PSTN Calls
In this task, you will configure digit manipulation to properly present the calling number and
transform the called number when receiving calls from the PSTN, so that the callback works
without editing the number at the phone.
Activity Procedure
Complete thesesteps:
Step 1 Manipulate the numbers in inbound PS'I'N calls on your IIQ gateway. Configure
appropriate translation prolilesand rules to meet Ihese needs:
Apply the correct prefix to the received calling number toenable callback:
Ifthe TON is subscriber, prefix the PSTN access code 0(local calls).
Ifthe ION isnational, prefix the PSfN access code 0and another 0 for
Luropean national calls.
If ION is international, prefixthe PS'I'N access code0 and 00 for
huropean international calls.
Convert thecalled number from the DID format (local, national, or
international) to an internally used extension.
Note Simulated PSTN delivers the entire called number, except for the leading zeros. This is
different from real PSTN, which delivers the local number after stripping the area and
country codes.
Apply the translation profile tothe ISDN PRI voice port in ihc incoming
direction.
Step 2 Manipulate the numbers in inbound PSTN calls on your BR gateway. Configure the
appropriate translation profiles and rules to meet these needs:
Apply thecorrect prefix lo thereceived calling number toenable callback:
IftheTON is subscriber, prefix fhe PS'I'N access code 9 (local calls).
11' the TON is national, prefix the PSTN access code 9 and I for North
American long distance national calls.
If the TON is international, prefix the PSTNaccess code 9 and Ol I for
North American international calls.
Convert the callednumber from the DIDfonnat (local, national, or
international) to an internally used extension.
Apply the translation profile tofhe ISDN PR! voice port inthe incoming
direction.
Step 3 Debugthe translations using the appropriate Cisco IOScommands.
You tested the translation rule by usingthe appropriate CiscoIOScommands.
42 Implementing Cisco Voice Communications and QoS (CVOICE) v8.0 2010 Cisco Systems Inc
You called your HQ phones from the PSTN phone (from various lines), using the correct
DID numbers (local, national, international), and verified that (he calls were delivered to
the desired endpoints and that the calling number was presented using the correct prefixes.
You called vour BR phone from the PSTN phone (from various lines), using the correct
DID number (local, national, international), and verified that the calls were delivered to the
desired endpoint and that the calling number was presented using the correct prefixes.
You enabled appropriate debugging on your gateways and monitored the translation
operations.
Task 5: Manipulate Calling Number and Called Number in
Intersite VoIP Calls
In this task, you will configure digit manipulation to enable intersite VoIP calls using intersite
prefix 8and these site codes: 10 for HQ and 20for BR.
Activity Procedure
Step 1 On your HQ gateway, enable HQ-to-BR VoIP calls using intersite prefix 8and these
site codes: 10 for HQ and 20 for BR.
Configure appropriate translation profiles and rules to meet these needs:
Add the prefix 810 tothe calling number in an outbound direction
towards the BR site.
Strip the prefix 810 from the called number in an inbound direction from
the BR site.
Modify the VoIP dial peer (3000) to match the complete number, including the
intersite prefixandsite code.
Apply the translation profile for an outbound direction to the VoIP dial peer
(3000).
Configure anew inbound dial peer (use tag 2) that matches all VoIP calls, and
apply the translation profile for the inbound direction to this new inbound VoIP
dial peer.
Atthe inbound VoIP dial peer (2). configure the same preferred codecs asthose
usedby the outbound VoIPdial peer (3000).
Step 2 On your BR gateway, enable BR-to-HQ Vol Pcalls using intersite prefix 8and these
site codes: 10 for HQ and 20 for BR.
Configure the appropriate translation profiles and rules tomeet these needs:
Add theprefix 820to thecalling number inanoutbound direction
towards the HQ site.
Strip the prefix 820 from the called number inan inbound direction from
the HQ site.
Modify the VoIP dial peer (2000) tomatch the complete number, including the
intersite prefix and site code.
Apply the translation profile for the outbound direction tothe VoIP dial peer
(2000).
Lab Guide 43
Step 3
Configure anew inbound dial peer (use tag 2) tliat matches all VoIP calls and
apply the translation profile for the inbound direction to this new inbound VoIP
dial peer.
At the inbound VoIP dial peer (2). configure the same preferred codecs as those
used by the outbound VoIP dial peer (2000).
Place atest call from the HQ site lo the BR site, using the intersite prefix and the site
code (dial 8203001 from an HQ phone). The call setup should succeed, bul you will
notice that, shortly after dialing, the caller ID is changed to the extension only
(3001). Ihis behav ior could confuse adialer. It can be disabled in both directions bv
using the following commands, entered at the HQ and BR gateways:
voice service voip
no supplementary-service h225-notify cid-update
You have completed this task when you attain this result:
You successfully placed intersite calls using the intersite prefix and site codes, and ><
verified that the calling and called numbers were presented correctly.
44 Implementing Crsco Voice Communications andQoS(CVOICE) v8.0
Lab 4-2: Implementing Path Selection
Activity Objective
In this activitv. vou will implement path selection lo ensure continuous VoIP service and toll
bypass. After'completing this activity, you will be able to meet these objectives:
Configure abackup PSTN path for intersite calls
Configure TEHO toprovide toll bypass
Visual Objective
The figure illustrates what you will accomplish in this activity. ^^^
Lab 4-2: Implementing Path Selection
Primary Path
Backup Path
TEHO to EU
TEHO to US
Required Resources
These are the resources and equipment that are required tocomplete this activity:
A PSTN phone
TwoCisco Unified IPphones in your HQ site
One CiscoUnified IPphoneinyour BRsite
Lab Guide
Command List
Digit Manipulation Commands
Command
digit-strip
prefix digits
forward-digits [0-32] | all | extra
num-exp dialed-digits substitution
voice translation-rule rule tag-
rule precedence /match/ /replace/
ftype fraatch-type replace-type} fplan
(match-plan replace-plan}]}
voice translation-profile profile-
name
translate {called | calling |
redirect-called} translation-rule-
number
translation-profile {incoming
outgoing} name
test voice translation-rule number
input-test-string [type match-type
[plan match-type]
debug isdn q931
debug voice translation
Description
Strips all thedigits thatexplicitly match the POTS
dialpeer. Digit stripping is enabled bydefault on
POTS dial peers.
Specifies the prefix ofthedialed digits fora dial
peer
Specifies which digitsto forward for voicecalls
Defines how to expanda telephone extension
number into a particular destination pattern
Defines a voice translation rule for voice calls
Defines a rule withina voice translation rule
Specifies a translation profile for all incoming VoIP
calls
Associates a translation rule with a voice
translation profile
Assigns a translation profile to a dial peer
Tests the functionality of a translation rule
Monitors ISDN Q931 signaling
Monitorstranslation operations
Job Aids
Ihesejobaids areavailable to help you complete the lab activil
Dial Plan
46
Thetable represents the dial plan that will be used inthe labs.
Site Internal Numbering Planand PSTN DID Ranges
HQ Site (EU)
BR Site (NA)
Internal extensions 2XXX
3XXX
Site codes 810
820
PSTN access code 0
9
Local DIDrange
555-2XXX
555-3XXX
National DID range 51p-555-2XXX
52p-555-3XXX
International DIDrange
55-51p-555-2XXX
66-52p-555-3XXX
Implementing Cisco Voice Communications andOoS(CVOICE) v8.0
Note
p 1 to 2 (pod number)
Calls from HQ(EU) to PSTN
Calls from BR (NA)to PSTN
Local calls
Example: 455-8000
Example: 455-8000
National
calls
O-NXX-NXX-XXXX, TON: unknown
Example; 0-455-455-8000
(3-digit area + 7 digits)
Example: 1-455-455-8000
International
calls
unknown
international
Example: 00-23-455-455-8000
011+ any number of digits, TON:
unknown
international
Example: 011-23-455-455-8000
Emergency
calls
112, TON: unknown
911, TON: unknown
Note
Nrepresents a digit between 2 and 9.
Task 1: Configure Backup PSTN Path for HQ to BR Calling
In this task, you will configure your HQ gateway for abackup PSTN path for calls from HQ to
BR.
Activity Procedure
Complete these steps on your HQ gateway:
Stepl
Step 2
Create a POTS dial peer (use tag3005) thatwill back upthe intersite calling toBR
viaPSTN when theVoIP path is unavailable. This POTS dial peershould have the
same destination pattern as, but a lower preference than, the VoIP dial peer.
Configure the digit manipulation mechanisms toensure that the call isdelivered to
the BRvia PSTN, andthat the callingnumberthat is presentedallowscallback.
Note Remember that HQand BRsites are virtually placed in different countries. Therefore, the
intersite that called numbers must have an international format for the call setup to succeed
via PSTN.
Step 3 Apply the digil manipulations tothe POTS dial peer tliat is used for the backup.
You have completed this task whenyou attainthese results:
Yousimulateda WAN failure by shuttingdownthe Serial 0/1/0interface on the HQ
gatewav.
Lab Guide
While the WAN was in the simulated failure, you placed acall to the BR phone as ifvou
were dialing via the WAN (the intersite prefix 8. site code 20, and 3001). The call should
have been redirected \ ia the PSTN.
The HQ toBR call was presented at the BR phone using anumber thai allows callback.
You tested the callback later, when the WAN was restored. You kept down the WAN fbr
the second task,
Ifvou attempted a rcserse call (BR toHQ). the call should have failed while the WAN was
in the simulated failure.
Task 2: Configure Backup PSTN Path for BR to HQ Calling
In this task, vou will configure your BR gateway for a backup PSTN path for calls from BR to
Activity Procedure
Complete these steps on vottr BR gateway:
Step 1 Create a POT Sdial peer (use tag 3005) that will back up Ihe intersite calling to BR
via PSTN when the VoIP path is unavailable. Ihis POTS dial peer should have the
same destination pattern as. but a lower preference than, the VoIP dial peer.
Step 2 Configure thedigitmanipulalion mechanisms loensure thatthecall is delivered to
the BR via PSTN, and thai the calling number that is presented allows callback.
Note Remember that HQ and BR sites are virtually placed in different countries. Therefore, the
intersite thatcalled numbers must have aninternational format for thecall setup tosucceed
via PSTN.
Step 3 AppK the digit manipulations lothe POTS dial peer that is used for the backup.
Step 4 Fix thetranslation rule(tag I) fortheprimary-ni issue, which wasdescribed earlier,
by adding a new rule that also sets the TON to unknown if the called-party number
starts with 011. This is required, because the current translation rule works fbr 9011
onlv. (The PSfNaccess code 9would normally beremoved by the dial peer 901.1
but thisdial peeris not used inthe PSTN backup scenario.)
You have completed this task when \ou attain these results:
While the WAN was inthe simulated failure (from previous lask), you placed a call loan
HQ phone as ifyou were dialing viatheWAN (theintersite prefix 8, sitecode l(), and
2001). The call should have been redirected via the PS'I'N.
The BR to HQ call waspresented at the HQ phone using a number that allowed callback.
You tested the callback later, when the WAN was restored.
You restored the WAN bv activating the Serial 0/1/0 interface.
Implementing Cisco Voice Communications andQoS(CVOICE) vS.O 2010Cisco Systems. Inc
Task 3: Configure TEHO at HQ for Calls to North America
In this task, you will configure TEHO for calls from the European HQ site to the North
American PSTN,
Activity Procedure
Step 1 On your HQ gateway, create aVoIP dial peer (use tag 6600) for tail-end hop off to
the North American PSTN via the BR gateway. The TEHO dial peer should be used
when calling to the virtual North American country with acountry code of66.
Remember to use the same codec class as for any other VoIP dial peer atthe HQ
gateway.
Step 2 Manipulate the sent calling and called numbers for the TEHO calling. The TEHO
configuration should not require any changes at the BR gateway. The calling number
that is presented should be acorrect international number that allows calling back to
the TEHO call originator.
When placing atest call from an HQ phone to the virtual North American country (for
instance. 66-455-455-I000#), thecallwas successfully routed viatheBRgateway.
You enabled appropriate debugging (ISDN, dial-peer) and verified that the expected TEHO
path was selected.
You verified that the correct calling number ininternational fonnat was presented on the
PSTN phone. The call should have rung up at the national, not the international, PSTN
phone button.
Task 4: Configure TEHO at BRfor Calls to Europe
In this task, you will configure TEHO for calls from the North American BR site to the
European PSTN.
Activity Procedure
Step 1 On your BR gateway, create aVoIP dial peer (use tag 5500) for tail-end hop off to
the European PSTN via the BR gateway, The TEHO dial peer should be used when
calling tothe virtual European country with a country code of55. Remember touse
the same codec class as for any other VoIP dial peer at the BRgateway.
Step 2 Manipulate the sent calling and called numbers for the TEHO calling. The TEHO
configuration should nol require any changes at the HQ gateway. The calling
number that ispresented should beacorrect international number that allows calling
back to the TEHO call originator.
Lab Guide 49
You have completed thistask when youattain these results:
When placing atest call from the BR phone to the virtual European country (for instance.
55-455-455-10()0#). the call was successfully routed via the HQ gateway.
You enabled appropriate debugging (ISDN, dial-peer) and verified that the expected TEHO
path was selected.
You verified that the correct calling number in international format was presented on the
PSTN phone. The call should have rung up at the national, not the international. PSTN
phone button.
50 Implementing Cisco VoiceCommunications and QoS (CVOICE) vS.O
Lab 4-3: Implementing Calling Privileges
Activity Objective
In this activitv. vou will implement calling privileges on agateway using COR. After
completing this'activity. you will be able to meet these objectives:
Create COR labels
Create COR lists and assign members
Assign COR lists to the appropriate dial peers and Cisco Unified Communications Manager
Express endpoints
Visual Objective
Lab 4-3: Implementing Calling Privileges
*k
V3 Phone i
2001
HQPhotnZ
2002
HQ-p
BR-o,
PODP
IP WAN
PSTN Phone-p
(Cisc UnifiedIP phone)
| P=pod number
Required Resources
A PSTN phone
TwoCisco Unified IPphones inyour HQsite
One CiscoUnified IP phoneinyour BRsite
)2010Cisco Systems, Inc.
Lab Guide 51
Command List
COR Commands
Command
Description
dial-peer cor custom
name cor-name
Specifies thatnamed CORs apply todial peers
Creates a named COR
Defines a COR list name
dial-peer cor list list-name
corlist incoming cor-list-name
Specifies the COR list tobeused when a specified dial peer
acts as the incoming dial peer
corlist outgoing cor-list-name
show dial-peer cor
Specifies theCOR list tobe used by outgoing dial peers
Displays COR labels
Job Aids
Ihese job aids are av ailable to help vou complete the lab activity.
Ihetabic defines call permissions asare required fortius lab.
Call Permissions
Endpoint Description
Permitted to Call
HQ phone 1
second line (2011)
Lobby
PSTN: Emergency
Internal: Any
HQ phone 2
first line (2002)
Executive (unrestricted) Any
HQ phone 2
second line (2012)
Sales
PSTN: Emergency, local, national
Internal: Any
BR phone
Employee
PSTM: Emergency, local
Internal: Any
All phones should bereachable from PS'I Nwithout any restrictions.
Note
TEHO destinations and PSTN backup for intersite WAN are not subject toCOR
configuration, and they can be reached without restrictions.
Site Internal Numbering Plan and PSTN DID Ranges
HQ Site (EU)
BR Site (NA)
3XXX
Site codes 810
820
PSTN access code 0 9
Local DID range 555-2XXX 555-3XXX
National DID range 51p-555-2XXX 52p-555-3XXX
International DIDrange
55-51p-555-2XXX 66-52p-555-3XXX
Implementing CiscoVoice Communications and QoS (CVOICE) v8.0
Note
p: 1to 2 (pod number)
Local calls
National
calls
International
calls
Emergency
calls
Calls from HQ(EU)to PSTN
Example: 455-8000
0-NXX-NXX-XXXX,TON: unknown
(0 +3-digit area +7 digits)
Example: 0-455-455-8000
unknown
international
Example: 00-23-455-455-8000
112, TON: unknown
Note
Nrepresentsa digit between2and 9.
Example: 455-flOOO
(1 +3-digit area +7 digits)
Example: 1-455-455-8000
unknown
international
Example: 011-23-455-455-8000
911, TON: unknown
Task 1: Configure Call Permissions for the HQ Site
In this task, you will configure COR names (labels), create COR lists with their members, and
assign CORlists to the appropriate dial peers and endpoints for call permissions that are
appliedto the HQsite.
Activity Procedure
Complete these stepsonyourHQgateway-
Step 1 Verify that unique dial peers exist for local calls, national calls, international calls.
and emergency calls.
Step 2 Createthe required CORnames.
Note Naming suggestions for COR names are: emergency, local, national, intl, lobby, executive,
and sales.
Step3 Createthe requiredCORlists.
Note Naming suggestions for incoming COR lists are: lobby-in, pstn-in, executive-in, and sales-in.
Naming suggestions for outgoing COR lists are: lobby-out, executive-out, sales-out,
emergency-out, local-out, national-out, and intl-out. ^^_
Step 4 Assign the COR lists to appropriate endpoints and dial peers in the correct direction.
Note
TEHO and PSTN backupdial peers are not subject to COR configuration.
Lab Guide
You verified that the call permissions, as they were defined in the job aids for the HQ site
were met by your configuration.
Task 2: Configure Call Permissions for the BRSite
In this task, you will configure COR names (labels), create COR lists with their members and
assign COR lists to the appropriate dial peers and endpoints fbr call permissions that are
applied to the BR site.
Activity Procedure
Complete these steps onyour BR gateway;
Step 1 Verify that unique dia! peers exist for local calls, national calls, international calls,
and emergency calls.
Step 2 Create the required number of COR names.
Step 3 Create the required COR lists.
Step 4 Assign the COR lists to appropriate endpoints and dial peers in the correct direction.
Note TEHO and PSTN backup dial peers are not subject to COR configuration.
You have completed this task when you attain fhis result:
You verified that the call permissions, as they were defined in the job aids for the BR site,
were met by your configuration.
Implementing Cisco Voice Communications and OoS (CVOICE) vS.O 2010 Cisco Systems Inc
Activity Objective
In this activitv vou will configure the corporate gatekeeper HQ as agatekeeper that controls
two /ones: 1IQ 'and BR. Call routing between the HQ and BR sites will be done via the
gatekeeper. After completing this activity, you will be able to meet these objectives:
Configure an H.323 gatekeeper to support multiple local zones
Register gateways at the gatekeeper
Configure technology prefixes
Implement CAC
Visual Objective
Required Resources
TwoCisco Unified IPphones inyour HQsite
One Cisco Unified IP phone in your BRsite
Command List
Lab Guide 55
Gatekeeper Commands
Command
gatekeeper
zone local zone-name domain-name
[ras-IP-address]
no shutdown
zone prefix gatekeeper-name el64-
prefix [blast J seq] [gw-priority
priority gw-alias [gw-alias, ...]]
gw-type-prefix type-prefix [hopoff
gkidl] [hopoff gkid2] [hopoff
gkidn] [seq | blast] [default-
technology] [gw ipaddr ipaddr
[port]]
bandwidth {interzone | total |
session} {default J zone zone-name}
.bandwidth-size
show gatekeeper calls
show gatekeeper status
show gatekeeper endpoints
show gatekeeper gw-type-prefix
show gatekeeper zone prefix [all]
show gatekeeper zone status
Gateway Commands
Command
gateway
h323-gateway voip id gatekeeper-id
{ipaddr ip-address [port]\
multicast} [priority priority]
h323-gateway voip h323-id
interface-id
h323-gateway voip tech-prefix
prefix
session target ras
Implementing Cisco Voice Communications and QoS(CVOICE) v8.0
Description
Enters gatekeeper configuration mode
Specifies a zonethat iscontrolled bya
gatekeeper
Brings the gatekeeper online
Adds a prefix to the gatekeeper zone list
Adds a technology prefix tothegatekeeper
configuration list
Specifies the maximum aggregate bandwidth
for H.323 traffic
Displays the status of each ongoingcall of
which a gatekeeper is aware
Displaysthe overallgatekeeper status,
including the authorization and authentication
status and zone status
Displays the status ofall registered endpoints
for a gatekeeper
Displays the gateway technologyprefix table
Displays the zone prefix table
Displays the status of zones that are related
to a gatekeeper
Description
Enters gateway configuration mode and
enables the gateway to register with a
gatekeeper
Identifies this as a VoIPgateway interface
Defines the name and location of the
gatekeeper for this gateway
Definesthe H.323name of the gateway,
identifying this gateway to its associated
gatekeeper
Defines the numbers that are used as the
technology prefixthat the gateway registers
with the gatekeeper
Enables RAS signaling, which means that a
gatekeeper is consulted to translate the
E.164 address into an IP address
Gateway and Gatekeeper Monitoring Commands
Command
Description
debug ras
Monitors Registration, Admission, and Status (RAS) messages
debug h225 asnl
Monitors H.225 ASN.1 library messages, which provide adetailed trace
of the RASmessages
Job Aids
Gatekeeper and Gateway Addressing
The table defines gatekeeper and gateway addressing.
HQSfte(EU)
BR Site (NA)
3XXX
Gatekeeper and Gateway Addressing
Component Address
Gateway H.323 ID
HQ gatekeeper Loopback0 IPaddress -
HQ gateway
Loopback 0 IPaddress
HQ-gw
BR gateway
Loopback 0 IP address -
Task 1: Configure Local Zones and Zone Prefixes
In this task, you will configure your HQ router as an H.323 gatekeeper that supports two local
zones. You will also configure zone prefixes toenable call routing.
Activity Procedure
Complete these steps onyour HQ router:
Step 1 On the HQ gateway, configure agatekeeper with these two local zones:
Local zone HQ, domain cisco.com, IP address ofloopback 0interface
Local zone BR, domain cisco.com
Step 2 Configure prefixes fbr the local zones IIQ and BR. Zone prefixes use the site
extensions as seen in the job aids section.
Step 3 Enable the gatekeeper process.
You viewed thegatekeeper status and verified thatthegatekeeper was up.
When viewing the gatekeeper zone status, you could identify the two local zones: HQ and
BR.
When viewing the gatekeeper zone prefixes, you could verify that they were correct.
Lab Guide
Task 2: Configure Gateways to Register with the Gatekeeper
In this task, you will configure the IIQ and BR gateways to register with the IIQ-based
gatekeeper.
Activity Procedure
Complete thesesteps:
Step 1 On your BR gateway, enable debugging ofRegistration, Admission, and Status
(RAS) messages.
Step 2 Configure >our BR gateway to register at the gatekeeper, using these parameters:
The H.323 interface should beLoopback 0.
H.323 bind using I oopback0.
Registering should be at zone BR.
Step 3 On jour BR gateway. configure anew Vol Pdial peer (use tag 2002) thai routes all
calls to the site HQ with extensions 2XXX to Ihe gatekeeper. Remember lo applv the
same voice-class codec asfbr olher VoIP dial peers.
Step 4 Configure your IIQ gateu ay to register al the gatekeeper using these parameters:
Interface hind using Loopback 0.
The 11,323 interface should beLoopback 0.
11.323 hind using Loopback 0.
The H.323 gateway IDis HQ-gw.
Disable the registration ofextension 2012 (second line ofHQ phone 2).
Registeringshould be at zone IIQ.
Step 5 On your HQ gateway, configure anew VoIP dial peer (use tag 3002) that routes all
calls to the site BR with extensions 3XXX to the gatekeeper. Remember to apply the
same voice-class codec as for olher VoIP dial peers.
Youhave completed this task when you attaintheseresults:
You were able tosee Ihe BR gateway registering at the galekeeper using ihe RAS
debugging and confirmed that the BR gateway had been registered.
You could confirm that the BR gateway had registered Ihe extension 3001 as F.164-ID.
You could confirm that the HQ gateway had been registered at the gatekeeper with the
correct 11.323 ID.
You could confirm that the HQ gateway had registered all extensions except 2012 as
F:.164-IDs.
When you placed an intersite call using an extension, the call succeeded. The calling
worked in both directions,
You examined the RAS registration and call admission messages using the debug ras and
debug h225 asnl commands, and you became familiar with Ihe H.323 debug output.
Implementing Cisco Voice Communications and QoS (CVOICE] v8.0 2010 Cisco Systems Inc
Task 3: Configure Call Admission Control
In this task, you will calculate the bandwidth requirements for one call and configure the zone
bandwidth for calls between theHQ and BR sites.
Activity Procedure
Step 1 Determine the codec that is used for intersite calling, and calculate the bandwidth
requirements for asingle call using the gatekeeper CAC calculation method. Write
these inthespace that is provided:
Note Remember that, actually, two different codecs can be negotiated between the gateways that
arebased onthe voice-class codec that isconfigured. For the calculation and CAC, you
must consider both codecs asthe initial call setup, though the gatekeeper will take into
account the worst codec of these two.
Step 2
Step 3
Configure the gatekeeper CAC to allow amaximum ofone call between the zones in
each direction.
Enable detailed RAS debugging, and observe that the second interzone call is
rejected at the gatekeeper.
You calculated that asingle iLBC call requires 30.4 kb/s (rounded up to 31 kb/s). using the
gatekeeper CAC calculalion method (2xcodec rate).
You calculated that asingle G.729 call requires 16 kb/s, using the gatekeeper CAC
calculation method.
You verified that one call can be placed successfully between the sites, while the second
call will fail.
Lab Guide
Lab 5-2: Implementing Cisco Unified Border
Element
Activity Objective
In this activity, you will implement fhe Cisco Unified Border Llcment features: protocol
interworking. various media flow methods, and codec transparency. Aller completing this
activ ity. you will be able tomeet these objectives:
Configure SIP-to-H.323 and H.323-to-H.323 protocol interworking
Implement codectransparent
Configure H.323-lo-ll.323 interworking
Implement media flow-around and media How-through
Visual Objective
The figure illustrates what youwill accomplish in thisactivitv.
Lab 5-2: Implementing Cisco Unified
Border Element
HO Phone t
2001
PODP
HQ Phone 2
2002
Required Resources
PSTN Phone-p
ICJsoi Unified IP phonsi
| P=pod number I
Ihese are the resources and equipment thai are required locomplete this activity:
Two Cisco Unified II* phones in your HQsite
One Cisco Unified IP phone in your BRsite
Implementing Cisco VoiceCommunications and QoS (CVOICE] v8 0
Command List
Cisco Unified Border Element-Related IOS Commands
Command
Description
voice service voip
allow-connections
Enables Cisco Unified Border Element protocol
interworking. Optionsare: H.323-to-H.323, SIP-
to-SIP, H.323-to-SIP, and SIP-to-H.323.
media flow-around | flow-through
Configuresmedia flow method. The default is
flow-through. This command is available in
voice service voip, dial peer, or voice class mode.
h323
Enters H323 configuration mode (fromvoice
service voip configuration mode)
eall start fast | slow |
interwork
Configures the H.323 signaling method. Default:
fast start.
codec transparent
Enables the dial peer to transparently pass the
codec proposals
codec
Configures a codec for a dial peer
show call active voice brief
Displays active call parameters, includingthe
RTP addresses
Displays brief informationabout active calls
show voip rtp connections
Displays RTP connections for active VoIP call
debug voip ipipgw
Monitors Cisco Unified Border Element
operations
debug ccsip messages
Monitors SIP messages
debug h225 events
Monitors H.225 events
debug h245 events
Debugs H.245 events
debug voip dialpeer
Monitorsthe matching of inbound and outbound
dial peers
Job Aids
Nojob aids are requiredto complete this labactivity.
Task 1: Configure Cisco Unified Border Element Functions
Inthis task, you will reconfigure yourdial plan toroute calls between your local HQ phones via
a BRgateway that works as anH.323-to-H.323 CiscoUnified Border F.lement.
Activity Procedure
Step1 On your HQgateway, configurea newVoIPdial peer (use lag 555) to sendall calls
to 5552XXX to the BR gateway:
Session target: Loopback 0 IP address of BR gateway
Codec: G.711 u-law
Step 2 On your HQ and BR gateways, modify the inbound VoIP dial peer (2) to accept only
the G.711 u-law codec.
Step 3 Onvour BR gateway, configure a newVoIP dial peer(usetag555)toreturn all calls
lo 5552XXX backto the HQgateway:
Session target: Loopback 0 IPaddress of IIQ gateway
Codec: G.711 u-law
Step4 Onyour IIQgateway, modifv the existingvoicetranslation rule that is associated
with the inbound VoIP dial peer, to remove 555 fromthe called number fbr the
inboundcall that is being returned back from BR.
Step 5 Trv to place a call between your HQ phones using the prefix 555 (for example, dia!
5552001 from IIQ phone 2). Yourcall will be blocked at the BRgateway, because it
lias nol been yet enabled for Cisco Unified Border Blemcnt functions.
Step 6 Lnablc Cisco Unified Border Element functions at the BR gateway toallow inbound
to outbound 5552XXX calling.
Step7 OnyourCisco Unified Border Element (BR), enable monitoring of Cisco Unitied
Border Element operations using the debug voipipipgw command, and place a call
between your HQphones, again using the prefix 555. This call should succeed.
Step 8 While the call isactive, issue tlie command showvoiprtp connections onyour
Cisco Unified BorderElement, to checkthe two call legsthat are associated withthe
call. Noticethat the remoteIPaddress is the same(IIQ gateway). Which codec is
used for this call? Write it in the space that is provided:
You have completed this task when vou attain these results:
You successfully placed an H.323-to-ll.323 call via the Cisco Unified Border Element and
observed how the call wasset up usingthe debug voip ipipgw command.
You could see the two VoIP call legs thai terminated at your Cisco Unified Border Element
when thecall wasactive. Ihe Cisco Unified Border Element used the default flow-through
method.
Task 2: Configure Codec Transparency
In thistask, vouwill configure Cisco Unified BorderElement not to participate incodec
negotiations.
Activity Procedure
Step 1 On >our HQ gateway, reapply the voice-class codec to (he VoIP dial peer (555) that
routes calls between HQ phones via the Cisco Unified Border Element.
Step 2 On vour HQ gatewav. reapplv the voice-class codec lo the inbound VoIP dial peer
(2).'
Step 3 Configure vour Cisco Unified Border Element BR to or participate in codec
negotiations for the calls between HQ phones at the inbound and outbound dial peers
(2 and 555). and accept an\ codec that is determined by the HQ gateway.
Step 4 Place a call between your HQ phones, using the prefix 555, and examine which
codec has now been used. Write the answer in the space that is provided:
You could see that the Cisco Unified Border Element did not actively participate in codec
negotiations and that it accepted the codec that was proposed by the HQ gateway. The HQ
gateway negotiates codecs that are based on the voice-class codec priority thai is
configured.
Task 3: Configure SIP-to-H.323 Interworking and Media Flows
In this task, you will reconfigure your Cisco Unified Border Element and HQ gateway for SIP-
to-H.323 interworking and modify the Cisco Unified Border Element to media flow-around.
Activity Procedure
Step 1 On your HQ gateway, modify the outbound VoIP dial peer (555) that routes calls
between HQ phones via the Cisco Unified Border Element, to SIP.
Step 2 On your Cisco Unified Border Element, modify the inbound VoIP dial peer (2) to
SIP.
Step 3 Enable your Cisco Unified Border Element for SIP-to-H.323 interworking functions.
Step 4 On your Cisco Unified Border Element, enable the debug ccsip messages, and place
a call between HQ phones using the prefix 555. Observe the SIP messages that are
setting up the call. Which SIP mechanism is used for the call setup, early or delayed
offer? Write the answer in the space that is provided:
Step 5 Configure your Cisco Unified Border Element for media flow-around. Place a call,
and use the command show voip rtp connections to examine that no RTP
connections are terminated at the Cisco Unified Border Element.
You reconfigured your Cisco Unified Border Element for SIP-to-II.323 interworking. and
the call via the Cisco Unified Border Element succeeded.
You observed SIP messages at the Cisco Unified Border Element as the call was setting up.
You could see that the early offer was used, since the SDP body was included in the SIP
INVITE message.
You could see that, if media flow-around was configured at the Cisco Unified Border
Element. RTP streams of an active call were not terminated at the Cisco Unified Border
Element.
) 2010 Cisco Systems, Inc Lab Guide 63
Lab 6-1: Implementing QoS Using Cisco
AutoQoS and Manual Configuration
Complete this lab activitv to practice what you learned in the related module.
Activity Objective
In this activitv. vou will implement fhe best-practice QoS mechanisms using Cisco AutoQoS
VoIP, and then \ou will manuallv tune the deployed QoS policy. Aller completing this activitv.
vou will be able to meet these objectives:
Configure AutoQoS VoIP on routers
Eine-tune the QoS policy on a switch
Verifv the operations of implemented QoS mechanisms
Visual Objective
The figure illustrates what vou will accomplish in this activity.
Lab 6-1: Using AutoQoS and Manual
Configuration
PODP
HO Phone 1
2001
HO Phone 2
2002
PSTN Phone-p
iGscuUnified IPpfto
P = pod number
Required Resources
64
These are the resources and equipment that arc required to complete this activity:
Two Cisco Unified IP phones in your [[Q site
One Cisco Unified IP phone in vour BR sile
Command List
Router QoS Commands
Command Description
clock rate
Configures the clock rate on a serial interface
bandwidth
Configures the available bandwidth on an
interface
auto qos voip Configures the AutoQoS VoIP feature. Available
in interface, subinterface, of Frame Relay DLCI
configuration mode.
priority Allocates specified bandwidth to the high-priority
queue
show auto qos
Displays the QoS policy that is deployed by
AutoQoS VoIP
show policy-map Displays the configured policy maps
Job Aids
No job aids are required to complete this lab activity.
Task 1: Evaluate VoIP Quality Without QoS Applied
In this task, you will examine VoIP quality when no QoS mechanisms are being used on the
routers.
Activity Procedure
Step 1 Ask the instructor to lower the clock rate on the WAN links to your HQ and BR
routers to 64.000.
Step 2 Simulate a heavy load by flooding the WAN using the ping command:
From the HQ gateway towards the BR gateway loopback (I IP address
Generate ping with 1000 packets, with a packet size of 5000 bytes
Syntax example (p is your pod number): ping 10.p.250.102 size 5000 repeat
1000
Note If you want fo interrupt the flood, simultaneously press Ctrl-Shift-6, followed by x.
Step 3 While the WAN link is congested, place an intersite direct call (that does not involve
the gatekeeper or Cisco Unified Border Element) by dialing while using a site code.
(For inslance, from an I IQ phone dial 820-3001, or in the opposite direction dial
810-2002.) Examine the audio quality in this way:
Speak into one receiver and listen to the sound al the other end. You should hear
a distinguishable delay.
On any participating phone, press the Settings button, select Status >Call
Statistics, and examine the parameters that are shown. The expected
approximate values are shown in the activity verification section.
2010 Cisco Systems, Inc Lab Guide 65
You have completed this lask when you attain these results:
You saw that Hooding the network with traffic caused perceptible delay.
You examined VoIP statistics and determined values in approximately these ranges:
A\erage jitter: 20-500 ms
Maximum jitter: 500-700 ms
Task 2: Configure AutoQoS VoIP
In this task, you will configure AutoQoS VoIP on the I IQ and BR gateways.
Activity Procedure
Step 1 On the HQ gateway, configure the correct bandwidth of 160 kb/s al the Serial
subinterface that represents the Frame Relay PVC to BR site (Serial0/l/0.l2l).
Step 2 On the BR gateway configure the correcl bandwidlh of 160kb/s at the Serial
subinterface that represents the Frame Relay PVC to FIQsite (Serial0/l/0.l 11).
Step 3 On the IIQ gateway. enter the interface-DLCI configuration mode for the DLCI 121.
and enable the AutoQoS VoIP feature while trusting the DSCP markers that are
received over the WAN.
Step 4 On the BR gateway, enter the interface-DLCI configuration mode for the DLCI 111,
and enable the AutoQoS VoIP feature while trusting the DSCP markers that are
reeehed over the WAN,
Step 5 Display the QoS policy that is generated with AutoQoS.
You examined the QoS policy that was deployed on the galeways when Cisco AutoQoS
was activated, using appropriate commands, including the show auto qos command.
Task 3: Fine-Tune QoS Policy
In this task, you will fine-tune the QoS policy that has been deployed by Ihe Cisco AutoQoS
VoIP feature.
Activity Procedure
Step 1 Calculate the bandwidthrequirement (including Layer 2) for a single VoIP eall,
using the follow ing formula:
BW= (codec payload + Layer 3+overhead + Layer 2 overhead) * packet rate *
8 bils per byte, with the following arguments:
G.729 payload: 20 bytes
Layer 3+ o\erhead (cR'I'P): 2 bytes
I aycr 2 overhead (FRF. 12): 8 bytes
Packet rate: 50 p/s
Implementing Cisco Voice Communications andQoS(CVOICE) vS.O 2010CiscoSystems, Inc.
Step 2 Examine how many calls are supported by the bandwidth that is allocated to the
LLQ that is provisioned by AutoQoS VoIP.
Step 3 On the HQ and BR gateways, reduce the LLQ bandwidth to support only two G.729
calls.
Step 4 Simulate a heavy load by flooding the WAN using the ping command:
From the HQ gateway towards Ihe BR gateway loopback 0 IP address
Generate ping with 1000 packets, with a packet size of 5000 bytes
Syntax example (p is your pod number): ping 10.p.250.102 size 5000 repeat
1000
Step 5 While the WAN link is congested, place an intersite direct call by dialing using a
site code. (For instance, from an HQ phone dial 820-3001, or in the opposite
direction dial 810-2002.) Examine the audio quality when QoS is implemented.
Compare the results with the Task 1 results.
You calculated that a single G.729 call with cRTP over FRF.12 requires 12 kb/s.
You calculated that two calls (24 kb/s) require 15 percent of a 160 kb/s total bandwidth.
You allocated 15percent of total bandwidth to the LLQand verified the setting using an
appropriate command, such as show policy map.
You evaluated VoIP quality with the same congestion in the WANand noticed an
impro\ ement in perceived voice quality due to shorter delay.
You examined VoIP statistics on the communicating phones and observed the values in
approximately these ranges:
Average jitter: 2 to 40 ms
Maximumjitter: 100 to 180 ms
2010CiscoSystems,lnc. LabGuide 67
Answer Key
The correct answers and expected solutions fbr the activilies that are described in this guide
appear here.
Lab 1-1 Answer Key: Configuring Voice Ports
Ihe HQ gateway configuration should be like the following:
Task l. Step I:
ip dhep pool HQl-Phones
! p is your pod number
network 10.p.2.0 255.255.255.0
default-router 10.p.2.101
option 150 ip 10.p.250.101
Task 2. Step I:
telephony-service
max-ephones 5
max-dn 20
ip source-address 10.p.250.101 port 2000
auto assign 1 to 2
Task 2. Step 2:
number 5552001
;
number 5552002
I
Task 3. Step 1:
r
Task 3. Step 2:
isdn switch-type primary-net5
lask 3. Step 3:
controller El 0/0/0
pri-group timeslots 1-8,16
f ask 4. Step 1:
destination-pattern 0[2-9]
port 0/0/0:15
nplementrng Cisco Voice Communications and QoS (CVOICE] v8 0 2010 Cisco Systems. Inc.
destination-pattern 00[2-9]
forward-digits 11
port 0/0/0:15
prefix 00
port 0/0/0:15
t
Task 4. Step 2:
port 0/0/0:15
forward-digits all
forward-digits 3
port 0/0/0:15
i
Task 5. Step 1:
direct-inward-dial
i
The BR gateway configurationshould be like the following:
Task 1. Step 2:
ip dhcp pool BRl-Phones
l p is your pod number
network 10.p.4.0 255.255.255.0
default-router 10.p.4.102
option 150 ip 10.p.250.102
i
Task 2. Step 4:
telephony-service
max-ephones 5
max-dn 10
ip source-address 10.p.250.102 port 2000
auto assign 1 to 1
i
Task 2. Step 5:
number 5553001
i
Task 3. Step 4;
2010 Cisco Systems, Inc. Lab Guide 69
isdn switch-type primary-ni
i
controller El 0/0/0
pri-group timeslots 1-8
i
Iask 6. Step I
destination-pattern 9[2-9 ]
port 0/0/0:15
destination-pattern 91[2-9 j..[2-9
port 0/0/0:15
prefix 1
destination-pattern 901IT
port 0/0/0:15
prefix 011
1ask 6. Step 2
forward-digits all
port 0/0/0:15
forward-digits 3
port 0/0/0:15
Task 7. Step I
direct-inward-dial
Lab 1-2 Answer Key: Configuring DSPs
The HQ gatewav configuration should include these commands:
Task I. Step 4:
ephone 1
codec ilbc
r
"lask2. Step 2:
voice-card 0
codec complexity medium
implemenling Cisco Voice Communications and OoS (CVOICE) vS.O 2010 Cisco Systems. Inc
Task 2. Step 6:
voice-card 0
codec complexity flex
i
Lab 2-1 Answer Key: Configuring VoIP Call Legs
The HQ gateway configuration should belike the following:
Task I, Step I:
'. p is your pod number
session target ipv4:10.p.250.102
no vad
i
Task 2. Step 1:
codec g723r53
i
Task 3. Step 1:
voice class codec 1
codec preference 1 g723r53
codec preference 2 ilbc
codec preference 3 g729br8
i
Task 3. Step 2:
no codec g723r53
voice-class codec 1
i
Task 4. Step 1:
no voice class codec 1
voice class codec 1
;
Task 4. Step 2:
voice-class codec 1
i
Task 5. Step 1:
voice service voip
h323
call start slow
Lab Guide
Task 5. Step 3:
interface LoopbackO
h323-gateway voip bind sreaddr 10.p.250.101
Task 6. Step 1:
i
Task 6. Step 2:
preference 1
I
Task 6. Step 4:
voice service voip
sip
bind all source-interface LoopbackO
i
Task 6 verification:
shutdown
The BR gatewav conllguration should be likethe following:
Task 3. Step 3:
voice class codec 1
codec preference 2 g723r53
j
l'ask 3. Slep 4:
voice-class codec 1
;
Task 4. Step 1:
no voice class codec 1
voice class codec 1
72 Implementing Csco Voice Communications and QoS (CVOICE) v8 0 2010 Cisco Systems, Inc.
Task 4. Step 2:
voice-class codec 1
;
Task 5. Step 3:
interface LoopbackO
h323-gateway voip bind sreaddr 10.p.250.102
I
Lab 3-1 Answer Key: Configuring Cisco Unified
Communications Manager Express to Support Endpoints
TheHQ gateway configuration should belike thefollowing:
Task I. Step I:
telephony-service
no auto assign 1 to 2
no auto-reg-ephone
Task L, Step 2:
no ephone 1
no ephone
7
2
Task 1. Step 3:
no ephone--dn 1
no ephone--dn 2
Task 2. Steps 2 to 4:
telephony-service
protocol mode dual-stack preference ipv4
cnf-file perphone
time-format 24
date-format dd-mm-yy
create cnf-files
I
Task 3. Step I:
number 5552001
i
number 5552002
t
Lab Guide 73
number 5552011
i
number 5552012
number 5552003
r
Task 3. Step 2:
ephone 1
mac-address 0024.c445.5233
type 7965
button 1:1 2:3
['ask 3. Step 3:
ephone 2
mac-address 0024.C445.4B7F
type 7965
button 1:2 2:4 3:5
i
Task 3. Step 5
telephony-service
create cnf-files
r
Task 4 (Optional). Step?
number 5552004
ephone 3
mac-address 0016.4155.B50B
type CIPC
button 1:6
j
The BR gateway configuration should include these commands:
Task l.Step 1:
telephony-service
no auto assign 1 to 1
no auto-reg-ephone
I
Task I. Step 2:
no ephone 1
Task I. Step 3:
no ephone-dn 1
74 Implementing Cisco Voice Communications and OoS (CVOICE) v8.0 2010 Cisco Systems, Inc.
1
Task 2. Step 5:
telephony-service
protocol mode dual-stack preference ipv4
cnf-file perphone
create cnf-files
i
Task 3. Step 6:
number 5553001
i
ephone 1
mac-address 0024.C445.4B48
type 7965
button 1:1
;
telephony-service
create cnf-files
i
Lab 4-1 Answer Key: Implementing Digit Manipulation
The HQ gateway configuration should be like the following (this isthe pod I configuration):
Task 2. Step 1
ephone-dn 1
number 2001
i
ephone-dn 2
number 2002
i
ephone-dn 3
number 2011
t
ephone-dn 4
number 2012
i
ephone-dn 5
number 2003
i
ephone 1
restart
j
ephone 2
restart
i
Task 3. Step I
2010 Cisco Systems. Inc . . _ ..
Lab Guide 75
rule 1 /A2/ /5552/
i
voice translation-profile pstn-out
translate calling 1
r
voice-port 0/0/0:15
Task4. Step I
rule 1 /.*/ IOf.1 type subscriber subscriber
rule 2 /.*/ /OOS/ type national national
rule 3 /.*/ /OOOf./ type international international
rule 1 /"5552/ 111
rule 2 /"5115552/ 121
rule 3 /"555115552/ 111
voice translation-profile pstn-in
translate calling 2
translate called 3
i
voice-port 0/0/0:15
lask 5. Step 1:
rule 1 /"2/ /8102/
i
voice translation-profile intersite-out
translate calling 4
i
translation-profile outgoing intersite-out
r
rule 1 /-8102/ 111
i
voice translation-profile intersite-in
translate called 5
translation-profile incoming intersite-in
76 Implementing Cisco Voice Communications and QoS (CVOICE) v8.0 2010 Cisco Systems, Inc.
voice-class codec 1
The BR gateway configuration should be like the following (this isthe pod I configuration]
Task 1. Step 3
rule 1 /'9011.*/ /&/ type any unknown
t
voice translation-profile fix-Oil
translate called 1
i
Task I, Step 4
voice-port 0/0/0:15
translation-profile outgoing fix-Oil
t
Task 2. Step 2:
ephone-dn 1
number 3001
i
ephone 1
restart
i
Task 3. Step 2:
rule 1 l'3l 155531
j
voice translation-profile pstn-out
translate called 1
translate calling 2
i
voice-port 0/0/0:15
i
Task 4. Step 2:
rule 1 /.*/ /9s/ type subscriber subscriber
rule 2 /.*/ I91&I type national national
rule 3 /.*/ /901W type international international
i
rule 1 /'5553/ 131
rule 2 /*5215553/ 131
rule 3 /"665215553/ 131
i
voice translation-profile pstn-in
translate calling 3
Lab Guide
translate called 4
i
voice-port 0/0/0:15
Task 5. Step 2:
rule 1 /"3/ /8203/
i
voice translation-profile intersite-out
translate calling 5
i
destination-pattern 810... .
translation-profile outgoing intersite-out
i
rule 1 /-8203/ 131
i
translate called 6
i
voice-class codec 1
I
Lab 4-2 Answer Key: Implementing Path Selection
The HQ gatewa> configuration should be like the following (this is the pod 1configuration):
Task I. Step 1:
preference 2
port 0/0/0:15
i
Task 1. Step 2:
rule 1 I'll /555115552/ type any international
i
rule 1 /"820/ /0066521555/
i
voice translation-profile pstn-backup
translate calling 6
translate called 7
Implementing Cisco Voice Communications and OoS (CVOICE) v8 0
Task 1. Step 3:
translation-profile outgoing pstn-backup
!
Task 3. Step 1:
voice-class codec 1
I
Task 3. Step 2:
rule 1 /"00066/ /91/
j
voice translation-profile teho-out
translate calling 6
translate called 8
I
translation-profile outgoing teho-out
i
TTie BR gateway configuration should be like the following (this is the pod 1 configuration):
Task 2. Step 1:
preference 2
port 0/0/0:15
i
Task 2. Step 2:
rule 1 /"3/ /665215553/ type any international
i
rule 1 /"810/ /01155511555/
i
voice translation-profile pstn-backup
translate calling 7
translate called 8
i
Task 2. Step 3:
translation-profile outgoing pstn-backup
i
Task 2. Step 4:
2010 Cisco Systems. Inc. Lab Guide 79
rule 1 /"9011.*/ /&/ type any unknown
rule 2 7*011.*/ /&/ type any unknown
i
Task 4. Step 1:
voice-class codec 1
I ask 4. Step 2:
rule 1 /'901155/ 1001
voice translation-profile teho-out
translate calling 7
translate called 9
translation-profile outgoing teho-out
;
Lab 4-3 Answer Key: Implementing Calling Privileges
The HQgateway configuration shouldinclude these commands:
Task 1. Step 2:
name emergency
name local
name national
name intl
i
Task L Step 3:
dial-peer cor list emergency-out
member emergency
i
dial-peer cor list local-out
member local
r
dial-peer cor list national-out
member national
dial-peer cor list intl-out
member intl
i
dial-peer cor list lobby
member emergency
Implementing Cisco Voice Communications and QoS (CVOICE) v80 2010 Cisco Systems. Inc
dial-peer cor list sales
member emergency
member local
member national
t
Task 1. Step 4:
corlist outgoing local-out
i
corlist outgoing national-out
i
corlist outgoing intl-out
i
corlist outgoing emergency-out
i
i
ephone-dn 3
number 2011
corlist incoming lobby
i
ephone-dn 4
number 2012
corlist incoming sales
r
The BR gateway configuration should include these commands:
Task 2. Step 2:
name emergency
name local
name block
i
Task 2. Step 3:
dial-peer cor list emergency-out
member emergency
i
dial-peer cor list local-out
member local
i
dial-peer cor list employee
member emergency
member local
>2010 Cisco Systems, Inc. ' '
Lab Guide
dial-peer cor list block
member block
i
Task 2. Step 4:
corlist outgoing local-out
i
corlist outgoing block
i
corlist outgoing block
i
j
ephone-dn 1
number 3001
corlist incoming employee
i
Lab 5-1 Answer Key: Implementing Gatekeepers
The UQ gateway configuration should be like ihe following (this is the pod Iconfiguration):
Task I. Steps I to 3:
gatekeeper
zone local HQ cisco.com 10.1.250.101
zone local BR cisco.com
zone prefix HQ 2...
zone prefix BR 3...
no shutdown
;
Task2. Step 4:
interface LoopbackO
ip address 10.1.250.101 255.255.2 55.255
h323-gateway voip id HQ ipaddr 10.1.250.101
h323-gateway voip h323-id HQ-gw
ephone-dn 4
number 2012 no-reg
gateway
.- ^^ c .r\rnirp* ft n 2010 Cisco Systems, Inc
nplementmg Cisco Voice Communications and QoS (CVOICE) v8 0
Task 2. Step 5:
session target ras
voice-class codec 1
i
Task 3. Step 2:
gatekeeper
bandwidth interzone zone HQ 31
bandwidth interzone zone BR 31
i
The BRgateway configuration should be likethe following (thisisthe pod I configuration):
Task 2. Step 2:
interface LoopbackO
ip address 10.1.250.102 255.255.255.255
h323-gateway voip id BR ipaddr 10.1.250.101 1719
i
gateway
i
Task 2. Step 3:
session target ras
voice-class codec 1
t
Lab 5-2 Answer Key: Implementing Cisco Unified Border
Element
The 1IQ gateway configuration should be like the following (this is the pod 1 configuration):
Task 1. Step 1:
codec g711ulaw
;
Task 1. Step 2:
no voice-class cl
codec $711ulaw
2010 Cisco Systems, Inc. Lab Guide 83
ask I. Step 4:
translate called 5
r
rule 1 /"8102/ 111
rule 2 /-5552/ IZl
;
fask2. Step 1:
voice-class codec 1
i
Task 2. Step 2:
voice-class codec 1
i
Task 3, Step I:
r
The BR gatewav configuration should he like the following (this is fhe pod I configuration):
Task I. Step 2:
no voice-class codec
codec g711ulaw
i
Task I. Step 3:
codec g711ulaw
lask 1. Step 6:
voice service voip
allow-connections h323 to h323
Task 2. Step 3:
codec transparent
r
codec transparent
Implementing Cisco Voice Communications andQoS(CVOICE) v8.0 2010CiscoSystems, Inc.
Task 3. Step 2:
i
Task 3. Step 3:
voice service voip
allow-connections sip to h323
j
Task 3. Step 5:
voice service voip
media flow-around
j
Lab 6-1 Answer Key: Implementing QoS Using CiscoAutoQoS
and Manual Configuration
You should configure your HQ gateway with these commands (this is the pod I configuration):
Task 2. Step I:
interface SerialO/1/0.121 point-to-point
description to BR-1
bandwidth 160
i
Task 2, Step 3:
interface SerialO/1/0.121
auto qos voip trust
t
Task 3. Step 3:
policy-map AutoQoS-Policy-Trust
class AutoQoS-VoIP-RTP-Trust
priority percent 15
r
The HQ gatewav configuration should be like the following (this is the pod Iconfiguration):
class-map match-any AutoQoS-VoIP-RTP-Trust
match ip dscp ef
class-map match-any AutoQoS-VoIP-Control-Trust
match ip dscp cs3
match ip dscp af31
i
priority percent 15
class AutoQoS-VoIP-Control-Trust
bandwidth percent 5
Lab Guide 85
class class-default
fair-queue
r
no ip address
no keepalive
i
interface SerialO/1/0.121 point-to-point
description to BR-1
bandwidth 160
ip address 10.1.6.101 255.255.255.0
class AutoQoS-FR-SeO/1/0-121
auto qos voip trust
i
map-class frame-relay AutoQoS-FR-SeO/1/0-121
frame_reiay cir 160000
frame-relay be 1600
frame-relay be 0
frame-relay fragment 80
service-policy output AutoQoS-Policy-Trust
i
You should configure vour BR gateway wilh these commands (this is the pod 1configuration):
Task 2. Step 2:
interface SerialO/1/0.Ill point-to-point
description to HQ-1
bandwidth 160
i
Task 2. Step 4:
interface SerialO/1/0.Ill
auto qos voip trust
I
Task 3. Step 3:
priority percent 15
The BR gatewav- configuration should be like the following (this is the pod 1configuration):
class-map match-any AutoQoS-VoIP-RTP-Trust
match ip dscp ef
class-map match-any AutoQoS-VoIP-Control-Trust
Z fn c ,r\imrci us il 2010 Cisco Systems, Inc.
Implementing Cisco Voice Communications and OoS (CVOICE) v8 0
match ip dscp cs3
match ip dscp af31
i
priority percent 15
class AutoQoS-VoIP-Control-Trust
bandwidth percent 5
class class-default
fair-queue
!
no ip address
no keepalive
I
interface SerialO/1/0.Ill point-to-point
description to HQ-1
bandwidth 160
ip address 10.1.6.102 255.255.255.0
class AutoQoS-FR-SeO/1/0-111
auto qos voip trust
i
map-class frame-relay AutoQoS-FR-SeO/1/0-111
frame-relay be 1600
frame-relay be 0
frame-relay fragment 80
service-policy output AutoQoS-Policy-Trust

Cvoice Vol 1-3

Caricato da

Informazioni sul documento

Copyright

Formati disponibili

Condividi questo documento

Condividi o incorpora il documento

Opzioni di condivisione

Hai trovato utile questo documento?

Questo contenuto è inappropriato?

Copyright:

Formati disponibili

Cvoice Vol 1-3

Caricato da

Copyright:

Formati disponibili

Table of Contents "

T1 PRI for Staridaracalis

Notification Requesl (RQNT)

Verify the received directory numbers

Cisco Baseline Mechanisms

Potrebbero piacerti anche