Prcparcd

CHAPTER

by-

Prof. lvld Saifur Rahman

10

CELLULAR MOBILE COMMUNICATIONS

Dr. Md, Saifur Rahman
Professor, Department of Elcctrical and Electronic Engincering

BtItlT. Dhaka

1000.

Mobile telecommunication is one of the fastest grorving and most demanding of all
telecommunications technologies. Currently, it represents an increasingly high percentage of
all nerv telephone subscriptions worldwide. In many cases, cellular solutions successfully
compete with traditional rvireline networks and cordless telephones. In the future, cellular
systems employing digital technology rvill become the universal method of
telecommunication.

History of Wireless Conmunication

In the early mobile radio systems, the desigr objective was to achieve a large coverage area

by using a single high-porvered transmitter with an antenna nounted on a tall tower. FM

mobile telephone system of the late 1940s used 120 KHz of RF bandwidth in a half duplex
mode (only one person on the tclephonc call could talk at a time), even though actual
telephone grade speech occupies only 3 KHz of baseband spectrum. The large R-F bandwidth
rvas used because of the difficulty in mass-producing tight RI filters and low noise, fiont-end
receiver amplifiers. With advent of improved technology, the FM channel bandrvidth rvas cut
to 30 KI-lz by the rnid i 9(rOs and thc spcctmnr e fficiency increased to four times rvith no nerv

spectrum allocation by FCC (Federal Communications Commission).

In

1950s and 1960s,

automatic channel trunking \yas introduced and implemented under the label IMTS
(lmproved Mobile Telephone Service) and with IMTS telephone companies.,began offering
full duplex, auto-dial, auto-trunking phone systems. Horvever, IMTS quickly became
saturated in major markets. By 1976, the Bell Mobile phone service for the New York City
market had only t',velve channels and could serve only 543 paying custonters. The
govemment regulatory agencies could not make new spectrum allocations in proportion to
the increasing demand for mobile services. So it becomes imperative to restructure lhe
radiotelephone system to achieve high capacity rvith large coverage area.

During the 1960s and 1970s, AT&T Bell laboiatories and other teiecommunications
companies throughout the rvorld developed the theory and techniques of cellular
radiotelephony - the concept of breaking a coverage zone into small cells, each of which

reuse the portions of the spectrum to increase the spectrum usage with $eater capacity at the
cxpense of greater system infrastructure. In 1983, the FCC allocated 666 duplex channels (40
MHz of spectmm in the 800 MHz band, each channel having a one-way bandwidth of 30 kHz
for a total specfmm occupancy of 60 kHz for each duplex channel) for the U. S. Advanced

N'lobile Phone System (AMPS). AMPS is an analog system that uses FDN1A to maximize
caDaclry.

The world's first cellular system was implemented by the Nippon Tglephone and Telegrjrph
conrplny (NTT).in.lapan. fhc system deployed in 1979, uses 600 FM duplex channels (25
one-"aj tink'; in the 800 MHz band. In Europe, the Nordic Mobile Telephone
Rfz-6r e-each
sysrem 0\i\{T 450) *,as clereloped in 1981 for the 450 MHz band ildZ5 kHz ch-anneK-Ifie

A"..ss

Cornnrunicatron Syslem

(t:lACIIof

and E!q$!-d-el$s!dqq fqsl

LIK rrc virturllv identicrl to the tlS AN{PS

154
(

i.

\,4

system but use 900 MHz band and smaller. bandtvidth channei. These :..::e::: ::: -:.,r bcine
replaced by
(Globrl S:.::-- ,:: l.irbile),
Dy the
tne Pan
Han European
tiuropean digital
dlgltal cellular standard 9Sltl
GJ]{ liGlobal
which was first deployed in 1990. A cellular system bas6d on code drrrs:tr-. :*]ll--3,s

(CDMA) has been developed

@I
supports variable

by

Qualcomm,

lnc. and sranaar:;:i :.--E

as an |JrLeriq

!4n!g4l!S,95, --:.,: i.. :::,--

number ofusers in 1,25 MHz wide channels using

dire-tsequ:::: ::::::

spectrum.

The origins of mobile communications follorved quickly behind the invention of radio j; i:.:
late 1800s. The first applications of mobile radio were related to the navigation and safeq c;
ships at sea. As radio concepts developed, so did it's use as a communications tool. The
major milestones in the development of rvireless communications are summarized in the

followine table.
able l: The Maior Milestones in the Developmeot of Wireless Communications
Date
Activity
r 906
Reginald Fesseden successfully transmits human voice over radio. Up until thal time,
radio communications consisted of transmissions of Morse Code
1915
J. A. Flemming invents the vacuum tube making it possible to build mobile radios.
t92t The Detroit Police Department (US) used a 2 MHz frequency in the departmenr's
first vehicular mobilc radio. The system was only one way and police had to find a
rvireline phone to respond to one way messages.
1930s Amplitude Modulation {AM\ two-way mobile systems werc in place in the U.S. that
took advantage of newly developed mobile transmitters and utilized a "Push-to-Talk"
or half-duplex transmission. By the cnd of the decadc, channel allocation gren' fronr

1lto40.
1935

lnvention of Frequency Modulation (FM) improved audio quality. FM eliminatcd the

1940s

need for large AM transmitters and resulted in radio equipment rvhich required less
power to operate. This made the use of transmitters in vehicles more practical.
The Federal Communications Commission (FCC) recognized a comnrunication
service it classified as Domestic Pubiic Land Mobile (DPLM) radio service. The first
DPI-M systcm rvas estffid
the 150 MHzband.

'the following year, a "highway" system was developed along'the Nerv York

1947
r

948

1949,

Boston corridor usins the 3 5 40 MHz band.

D. H. Ring, working at Bell Laboratories, envisions the cellular concept.
Shockley, Bardeen and Brittain, at Bell Laboratones, invent the transistor rvhich
enables electronic equipment, including the radio to be miniatr.rrized.
Bell Systems (US) made broadband proposals

1958

t964
1968
1

969

AT&T introduces Improved Mobile Telephone System (IMTS)

The FCC began to address issue of nerv US spectrum requirements.
Nordic countries of Denmark, Finland, Iceland, Norway and Sweden agree to form a
group to study and recommend areas of cooperation in telecommunication. This led
to the standardization of telecommunications for all members of the Nordic Mobile
Telephone (NMT) group, the first comprehensive Intemational standardization
group.

The NMT g-oup

T! u=:. ne8glb.
Th e FC
C

authorize

fh'is {.utut. *oqld becJme tt e uu.ls oi.ourn,nn

trr@H,"r

system in the US (Illinois Bell Telephone Comna-n,r)

Ericsson launche, th"

*a

*;i;a;m

*o

analog NMT 450 standard.

cellular standard (GS

Th. nu-b.. of rnobil" rub

Limitations of conventional mobile telephone systems

one of many reasons for developing a cerurar mobile
telephone system and deploying it in
many cities is the gperglional li{tations of convenrional
,Jbil"-t"t"phon"

.
.'

,yrt.n.,,

Limited service capabilitv

Pjgllgrvice nerformat'ce
And inefficient lrequency spectrum utilization.

Nlobile Standards
Standards play a major role in telecommunications
bv

o 4$_i!g uroducts flom diverse suppliers to ieinterconnected.

. LasiljJatils
innouation by creating la.se ma.ke1!ts?Eifiii?odu"ts.

The standards-making process

o1e of co_operation at many levels, both nationally
and
1s
intemationally and includes cooperatton
bchvecn:

o
o
o

Industrial concems wilhin a country

These industrial concems and their govemments
National govemments at an intemational level.

The primary purpose of a standard for mobile communications

ls to specify how mobile
phone calls are to be handled by a mobile n.t*o.t.
for exumf,ie, tnrs includes specification

ol the lbllowing:

.
e
o
.
.

The signals to be transmitted and received by the mobile phone

The format ofthese signals

The interaction ofnetwork nodes

The basic network services which should be available
to mobile subscribers
The basic network structure (i.e. cells, etc.).

Since the development of NMT 450 in r 9g r many standards

,
for mob e communication have
been developed throughout the rvorld. Each nlobiie standard
has been developed to meet the
particular requirements of the country- or the ihterest group,
inuotu"o n its specification. For
this reason, although a standard may be suitable ro,
6n"
rt may no1 be suitabre for
another' The main standards and the main markets in which
"ountry,
they are used are summarized in
the followins table.

156

PrcparcJ b1

Prot-.

i\ld Srifur

Rahman

rVain Stald
ain Nlarkets of Cellular llobilc Comnrunicatiou
Year Standard
Nlobile Telephona Svstenr
Tcc hnologr
Printar)'NIarkets
I 981
NMT 450 r\orolc t\40blle I eleDhonv
Analog
Eurgpe, lt tidd Ie Erst
l98l AMPS
Aovanceo NloDlte Phone system
Analog
North and
Iable,2_i

Sourh

198s

Total Access

TACS

Amenca
Communicatton

Anaiog

Europe and China

Analog

Europe, Middle East

System
986

NMT

t99t

GSM

Nordic Molile Telephony

9OO

Global System for

Mobile

Communication

i 991

1992

D.AMPS

Digiral-AMpS

1800

World-rvide

North and

Digital

Global System for-

GSM

Digital

Mobrle

Digital

Europe

Digital
Digital

l.{orth Anterica

Communication

1994

Dna

1995

LC!900

South

Amenca

Ielsenal Digital Cellular

Pe rsojqlCor:rmuniquionlgrviccs

Japan

History of GSNI: The history of GSM is outlined in the following tabie:

Table 3; The llisto
Date

t9821985

oOf GSNI

Activity
..:-

S:":lfiT:e;|llopi"
!@in thc 900 Muz frcqucncy blnrr. Tiis
drgrlal rcrccomrrunrcations sta'dard

E'uropean
standard later became known o. Gl:bnl

1986

1987

1988

1989

SI:r!.,,fT yo

n technolosy to use.

A Combination of TDMA

una

noun@

for GSM
Operators from l2 countries signed a M
themselves to introducing GSM bv 1991 .
CEPT began prodr.rcing CSM s
five countries signed the MoU.

Il.,-^-

^--

f-1-

UJLU

SDeCI I

MoU

tClrtlon.

1990

cqurnment.
I 991

The GSM i 800 standard

1992

@wingoutside
Phase I specifications were completed.

CEPT to siga.

First commercirl Phase I GSM netlorks rvere launched.

The first Intemational roaming agreement was established between Telecom
Finland
and Vodafone in UK.
1993

The MoU norv had a total of 70 signatories, GSM networks rvere launched in

t994

Norway, Austria, Ireland, Hong Kong and Australia.

The number of GSM subscribers reached I million.
The first coqmercial DCS 1800 system ivas launched in the UK.
T'
l ne tvloU now nf,d over lU0 slgnatories covering 60 countries.
Nlore GSM networks uere hunched.
The tgtal number of GSM subscribers exceecled 3 miilion.

frcp.rrcJ bv - Proi l\ld S.ulirr R.rhntrn

ii* .f.tf'-tt"" f"t the Personal Communications

Services (PCS) was developed in

the USA. This version of GSIvI operates at i900MHz.

Th" ft.rt CI.SM'1900 Sy"t".. became available These comply rvith the PCS 1900

At th. b"gr""Gg

tgqg the \4oU has a total of 253 members in over 100 countries

"f
and there are over 70 million GSM subscribers world-wide. GSM subscribers account
for 31% of the world's mobile market.

Because GSM provides a common standard, cellular subscribers can use their telephonqs
over the entire GSM service area which includes all the countries around the world rvhere the
GSM system is used.
as high sPeed data communication, facsimile
and a Short Message Service (SMS). The GSM technlcal specltlcatlons ale also desrgneo to
work with other standards as it guarantees standard interfaces.

ln addition, GSM provides user services such

Finally, a key aspect of GSM is that the specifications are open ended and can be built upon
to meet fuhrre requirements
GSM Specifications and GSM Phases
GSM rvas designed to be platfom-independent. The GSM specifications do not specif-v the
actual hardware requirements, but instead specifu the netwolk functions and interfaces to be
creative in how they provide the actual functionalify, but at the same time makes it possible
for operators to buy equipment from different suppliers
The GSM recommendations consist of twelve series which are listed in the table below.
groups A
These series rvere written by different working parties and a number of expert
permanent nucleus was established in order to coordinate the working parties and to manlge
ihe editine of the recommendations. All these groups were organized by ETSI'
Table 4: GSM Recommendations
Content

Series
01

General

0z

Sewice aspects
Network aspects
MS-BSS interface and Protocol
Physical layer on the ra4io Path
Speech coding speci!gq!9!
Terminal adapter for MS
BSS- MSC interface
Network interworking
Service interworking
Equipment and ry?e app1qyql JP9!ifi99!9n

03

04
05
06
01
08

09

l0
1l
t2

Operation and maintenance

158

P:el::ei :r - prof

l'hc GSM 1800 section ,, *:,111,1ru^l"_ll"^ryn *,,n,n

* *,

lld

Sarfur Rahnran

."*,nrn**or.s.

',,
"i? has
oo o,n, e00 s based on
iT';"11i*::::,:
been
i:y::t.f:*j9d;,,
iffi
i""'ii',il."il#.:1".:ifiiiglf:::l::lJ1"r?1",$J;
"a"pi.a
:"'#"li:'and

e:Ti

[1%

GSM Phases

l;'hlil:,

J.1?:;il',ffiTi,l:":il".1d:,:"#::ml'.".1

::jxn.*:,'Jdnfr

;j,*,tru:*ff

;;ffi
::,:ffJ ffi::1 ".,

j:i1.,

*li,,l;:U*
jtt."ixl;T#

J1i:it*l

Phsse 2

1989

E
Phase

Phase

Slandardizalion

ldplehsnrario^/t/sa9.

Figure: CSM phases

contains the most conlmon

services including:

Voice telephony
Internationalroaming
Basic faxldata services (p to
9.6 kbits/s)
Call forwarding
Call barring

.'
.
r
o
o

Short Message Service (SMS)

Phase I also incorporated

featr
cards phase , ,o.ii"",r"". *lii'.;::i,:::f*':ffix1;JH::,r.:Jr1dentiry Modure (SrM)
Phase 2

Additional features were introd uced

in GSM phase 2 including:
o Advice ofcharge
. Calling line identification
o Call rvaiting
o Call hold
o Conference calling

r
o

Closed user groups

Additionaldatacommunicationscapabilities

Phase 2+

The standardization groups have

already begun to define the
nr
programs *irr

.ou", iurirpre Jubscriber numbers

159

and a

*n.,,

"ii,rf,X.rr?;:"ffilffii-

Prepared by

- Prof Nld Saifur

Rahman

Some of the enhancements offered by Phase 2+ include:

r
o
.
o

Multiple sewice profiles

Pri.,rate numbering plans

Access to Centrex services

Interworking with GS 1800, GSM lg00 and the Digital Enhanced cordless

Telecommunication (DECT) standard.

Priorities and the time schedules for new features and functions depend primarily on the
interest shown by operating companies and manufacturers and technical develoDments in

related areas.

GSM Network Components

The GSM network is divided into two systems. Each of these systems is comprised of a
number of functional units which are individual components of the mobile network. Thje two
systems are:

o
o

In

Switching System (SS)

Base Station System (BSS)

addition, as with

all

telecommunications networks, GSM nehvorks are operated,

maintained and managed from computerized centres.

il
L:;--2

Switching System

l:Rff

lAU.o l

--,ugt

Base Station System

J
/'q\

- - - - Signaling kansmission
Call connections and
signaling transmission

,,,

/
\*
Figure: GSM Slstem model

Abbreviations:

AUC - AUthentication Centre

BSC - Base Station Controller
BTS - Base Transceiver Station
EIR - Equipment Identity Register
HLR - Home Location Register
MS - Mobile Station

160

J
r:
l-

l)rcplrcd b\'-. Prol

\ld

Slrtirr

Rahm.rn

I\ISC - Nlobile sen ices Switching Centre

NMC Network Management Centre
ONIC - Operation and Nlaintenance Center
\.LR - Visitor Location Register
The SS is responsible for performing cali processing and subscriber related functions.
The
BSS performs all the radio-related functions. The olIC performs all the operation and
maintenance tasks for the network such as monitoring nenvork traffic and network alarms.
The oMC has access to both the SS and the BSS. MSs do not belong to any of these sysrems.

GSM Geographical Network Structure

Every telephone nehvork needs a specific struc re to route incoming ca s to the conect
exchange and then on to the subscriber. In a mobile nenvork, this structure is very important

because

the subscribers are mobile. As subscribers move through the network,

structures are used to monitor their location.

these

CELL
A cell is the basic unit of a cellular system and is defined as the area of radio coverage given
by one BS antenna system. Each cell is assigned a unique number called cell Global ldJnriry
(cGI). In a complete network covering an entire country, the number of cells can be quile

hieh.

/r)

l-*
t\,A

/q\'&

'\

Figure: A cell

LOCATION AREA (LA)

A Location Area (LA) is defined

as a group of cells. within the netrvork, a subscriber's

location is knorvn by the LA which they are in. The identity of the LA in which an MS is
currently located is stored in the VLR as Location Area Identity (LAI).
When an MS crosses a boundary from a cell belonging to one LA into a cell bclonging to
another LA, it must report its new location to the network. when an MS crosses a iell
boundary within an LA, it does need to report its new location to the nehvork. when there is
a call for an MS, a paging mcssage is broadcast within all cells belonging to an LA.
NISC SERVICE AREA

An MSC service area is made up of a number of Las and represents the geographical part of
the netrvork controlled by one MSC. In order to be able to route a call to an NiS, the
sr.rbscriber's MSC sen'ice area is also recorded and monitored. The subscriber's lv{SC senice
area is stored in the FILR.

Prco.rrcrl

[rv I'rof \1J Srrtirr RuhmlI

MSC scrvice arca

PI-}I}-.

IJE

iI.VICE ARiiA

A Public Land Mobile Nehvork (PLN{N) senice

area is the entire set of celis sened by one

llet\\ork operator and is dcfined as the area in rvhich an operator offers radio coverage and
access to its netrvork. In anl,one countn there may bc sevcral PLNIN scn'ice areas, onc for
each mobile opcrator's nenvork.

csll sEtrvtcIl ,\Rli,\

I-he GSM service area is the entirc geographical arca in which a subscriber can gain access to

US\1 nctrrork. lhc ('sl\l srrri.c rrrer inercirsc.' iis nlorc opr'rit(or:\ "iFn contrr(t5 ltrcclng
t() work togcthcr. CLrrrcntly, thc GSlvl scrvicc arca spirns dozcns of counlries lcross thc *'orld
from Ircland to Australia and South Africa

intcmational roaming is thc telr| applied rvircn an

N'IS

uoves from one PLNln" to another.

Location A

@
[:igLrrc Rclation bciwccn arcas in CS]\l

162.

Prcpared

b;

Prof.

i\ld

Sarfirr Rahman

GSM FREQUENC]Y BANDS

As GSM has grown worldwide, it has expanded to operate at three frequency

bands: 900,
I ti00 and 1900 MHz.
GSM
GHz

0.8

9OO

GSM

1.2 1.3 1.4 1.5

0.9

t800

GSM 19oO

1.6

Figure . GSM frequency bands.

GSM

9OO

GSNI

18OO

In 1990, in orcler to increase conrletition hetwen operstors, the llnitcd Kingdom

requested
the start of a new version of GSM adapted to the l g00 MHz frequency b-and-.

tE;;;

ha!.;

l::l'i'::i,:il""1","'.*:s'i:"1i0"1""Y-"'5fl
:ll.turroperationlli'-"'qdeqslui&;
Digital Cellular Syste4 (Q9!_)_l!!gJ4IIg. In 1997
it was ienamed cSM I 800.
By granting licenses lor GSM 1800 in addition to GSM 900, a counrry can
rncrease the
number of operators. In this way, due to increased competition, the
serviie to subscribers
is

rmproved.

GSN| 1900

In

1995, the P:.rgr.lal

oncept was specified in the United
,,station-to5tates. l he basrc ldea ls to enable "person-to-person" communication rather
than
station". PCS does not require that such services be implemented using cellurar technologry,
but this has proven to be the most effective method. Tire frequencies availabre for pcS
are
around 1900 MI-lz.

9"rr""t.

KEY TERMS
During the development of mobile systems, many terms arose which are used
to describe the
call cases and situations involving MSs. The primary terms used are described below.
Ann MS can have one of the follorving states:

Idle: the MS is ON but a call is not in progress.

Active : the MS is ON and a call is itt progress
Derached: the MS is OFF.

I'rcpllctl

bl

Prof i\ld Slrlirr

R.rhnrun

Tlre foilorving table defines the key terms used to describe GSM mobile traffic cases (There

are no traffic cases in detached mode).

Table
Mode

Idle

erms fo define GSM Mobile Traffic Cases

Description
Registra tion
This is the process in which an MS
Term

attached

Roaming

Location
Updating
Paging

Activc

I'landover

-ifrat

itG

When an MS moves around a network in idle mode, it is refer"d !o

as

Intemational
Roaming

informii nEtwo

loamlns.

When an MS moves into a network which is not his home network

it is refened to as intemational roaming. MSs can only roam into
networks with which the home network has a roaming agreement.
An MS roaming around the network must inform the netrvork rvhen
rt enters a nerv LA. This is cailed location updating.
This is the process whereby a nehvork attempts to contact a
particular MS. This is achieved by broadcasting a paging message
containing the identity of tbat MS.
This is the process irr which control of o call is passcd ftonr onea.cll
to another while the MS moves between cells.

NIS Registration and Roaming

when an MS is powered off, it is detached from the network. when the subscriber srvitches
power oN, the MS scans the GSM fiequencies for special channels called control channels.
When it finds a control channel, the MS measures the signal strength it receives on that
channel and records it. When all control channels have been measured, the MS tunes to the
strongest one.

When the MS has just been powered ON, the MS mr.tst register rvith the netrvork rvhich rvill
then update the MS's status to idle. If the location of the MS is noticed to be different fiom
the cunently stored location then a location update will also take place.

As the MS moves through the netlvork, it continues to scan the control channels to ensure
that it is tuned to the strongest possible channel. If the MS finds one which is stronger, then
the MS returns to this nerv control channel. Q.lote: In idle mode, it is the MS rvhich decides
lrhich cell to move into. In active mode, the network makes this decision.) If the nerv
control channel belongs to a nerv LA, the MS will also inform the network of its new
location.

I64

Prepared by

Prof. lvld. Saifur Rrhman

@A

Figure:Roaming

NETWORK COMPONENTS
It has the following components or sub-systems.

o
.
.
.

Switching System (SS) Components

Base Station System (BSS) Components
Network Monitoring Centres
Mobile Station (MS)

Slvitching System (SS) Components

Mobile services Switching Centre/ Visitor Location Register

(MS,l/LR)
NISC

An MSC performs the telephony switching switching functions for the mobile net\r,ork. It
controls calls to and from other telephony and data systems, such as the Public Switched
Telephone Network (PSTN), Integrated Services Digital Network (ISDN), public data
networks, private networks and other mobile networks.

An MSC can become a Gateway MSC (GMS) by integration of the appropriate software.
This enables an MSC to interrogate a network's HLR in order to route a call to a mobile
station (MS). For example,

if

a person connected to the PSTN wants to make a call to a GSN'I

mobile subscriber, then the PSTN exchange will access the GSM netrvork by first connecting
the call to a GMSC. The same is true of a call from an MS to another MS.

vLR
The VLR database contains information about all the mobile subscribers currently located in
an MSC service area. Thus, there is one VLR for each MSC in a netrvork. The VLR
temporarily stores subscription information so that the MSC can provide service to all the
165

Prcpared

b!

- Prol \1'l S:rlur R:h;rsa

regarded as
area The VLR can be
service
MSC
that
vistttng .,r-^ rrr D i.fnrnreii.n stored lbout
subscribcr'
---rbout the s.bscribcr'
subscribers currently

---

--t:"
ll,t,,.J,:til#il:,lirJillli"iTl,'lli*i*i"i"*"ii"n
rnto
when a subscriber roams

a,newi4SC.:lll$,i,il1;.i;:."riJil?H?tl::il::T"t;

:.:ill"'fi'ff :i:iHiq,l+'::::i::l*;'**1}:'H:;;::'xilT1ll:;:JJ
ff
ror can set-up'
fl;rumll'l"",X"""iti::',':;'#nl[*1i:*;'i,':?H:f,i,]l1'"asubscribcr

:i,t*'l|"ff ii':l,i".'ff ffiJ;ffiit'"

inro""ution'"quired

stores
nenvork database that
The HLR is a centralized

":9

T:f?:; "tt-'"btl:tt::;:lj;;:;

iix;uiffi l':hT:T,[:,1;'s"i:':.'P'"ffi:il"i'"'i"'"*oincrudes:
itffi
.
Subscriber identitY
servtces
Subscriber supplementary
informatton
Subscriber location

.
.
. iuit.tiu"t authentication informatton

:*l'fitl:

:n#il;ilJillifT:Ti'l,:T.::[""1"'y'T]'l':ff "?:li:#"'"'#:'x"ff
maY be added'

Itj"t""tkt"g L"t"ttt
(rLR)
rhe rnterworking Lorction.register

'*:*f

,A

il"hi*"

:,"i

:iy.,#.,il1""-"

(Auc) atd
Authentication centre

exists

ii)lJL]'"lY;^l
J c s *,
i,'ti.',
iill,: ; ,'T',ff
" "
1

SY'?l.i3

l!tii};;
\'
0

E@

AUC

anemtii8

l::tffi""::T:,jj}:i
;;i;
$l,j ;l$*.:+",:'"u: t";rr:"i:il'"""i:Hll"#'ff

the sub.scriber
rs to authenticate
The funcrion of rhe AUC

;nerin g kevs use'l'lo

ensure network secunty'

EIR

$:n:T;':"",*:::[''j#]1:t5:1;Jig:iJ"i1l,''.tr[{*Ilflil:"llit
'"::#:::j#Hilll'"
"K"no*

lnterface (DTI)
Data Transmission

the Drr

which imprcments
is a hrTdwsre pratfomr

lT':i,y"::T:1J5{i:-il'i':'T'L:""'"..:
qXE

!'Jl*,flil*::
:H il?"{ilt*$t';rJ*;1"'r"':"t
and m
operation
DTI does not needieparate
olrtform. the

ir"r. so.ion svstttJg!9lCsmps!g$l-I66

in'lo

',:n*t

'lhe

Prrp,rrcrl br,-

l'r,rl i\1rl S.rltrr

Il.rhrrt.tn

Tr0nsc oder Controller (TR C )

'fhc purpose of a TI{C is to mLlltiplex network tralfic channels from multiple BSCs onto one
6.{ Kbits/s PCM channel which reduces network transmission costs. The TRC can be
combined with the BSC or exist as a stand-alon node.
Base Station Controller (BSC)

'Ihc BSC manages all the radio-related frinctions of a GSM network. It is a high capaciry
su'itch that provides functions such as MS handover, radio channel assignntent and the
collection ofcell configuration data. A number ofBSCs may be conkoiled by each NISC.
Base Transceiver Station (BTS)

The BTS confiols the radio interface of the MS. The BTS comprises the radio equipment
such as transceivers and antenna which are needed to sewe each cell in the netrvork. A group
of BTSs are controlled by a BSC.
Network Monitorins Centres
Operation and Maintcnanc( C.'ntre

(oMrl

An OMC is a computerized monitoring centre which is connccted to othcr netl

ork

components such as MSCS and BSCs via X.25 data nehvork links. [n the OMC, staff are
presented with information about the status of the nefwork and can nronitor and control a
variety of system parameters. There may be one or several OMCs rvithin a nenl ork
depending on the network size.
Network Management Cenfi'e (NMC)
Centralized control of a network is done at a Network Management Centre Q'{MC). Only one
NMC is required for a network and this controls the subordinate OMCs The advantage of
this hierarchical approach is that staff at the NMC can concentrate on long term system-wide
issues, rvheras local personnel at each OMC can concentrate on short term, regional issucs

OMC and NMC functionality can be combined in the same physical network node or
implemented at different locations.

Nlobile Station (MS)

An lvls is used by a mobile subscriber to communicate with the mobile netrvork.

Several

tlpes of MSs exist, each allorving the subscriber to make and receive calls. Manufacturers of
MSs offer a vanety of designs and features to meet the needs of different markets.
The range or coverage area of an MS depends on the output power of thc N{S. Diffcrcnt tlpcs
of MSs have different output power capabilities and conscquently diffcrent rangcs. For
example, hand-held MSs have a lower output powel and shorter lange than car-installed MSs
with a roof-mounted antenna.

t67

Prepared bv

Prof Md Saifur Rahman

Figure: Ranges ofdifferent types ofMS (Hand-hld MS and car-installed MS)

GSM MSs consist of:

A mobile terminal

o
r

A Subscriber Identity Module (SIM)

Unlike other standards, in GSM the subscriber is separated from the mobile terminal. Each
subscriber's information is stored as a "smart card" SIM. The SIM can be plugged into any
GSM mobile terminal. This brings the advantages of security and portability for subscribers.
For example, subscriber A's mobile terminal might have been stolen. However, subscriber
A's own SIM can be used in another person's mobile terminal and the calls wil'l be charged to
subscriber A.

168

i4

- Prof Nld. S3r!ur Rlhman

lt.3

i60 samples o!er 20

rns and l3bits/ sarnplc

ir'

b,s

r,

trits lrirnlc

Mic

I
8KHz

(20 ms speech

Sampling

segment)

260 bits/frame
and l3kb/s

l3bits/sample
( t 3Bir PCM)

456

I'r frame

G]V{SK

(20ms)
I

l,l

MOD

bits,/ tiame x 8 frames

456

33.85 kb/s per user

2nd frame

(20ms)

1,

Rh

= 8000 samplc/s

ll

T, = l/8000 = 0.125 ms

20

20000
No of sample
20
0.125 l2i

ms

hirs _

ll

birs .. $000 s .. .,, _..

sample -semple

^^

= 160 sample over 20 ms

= 270.3

fru.e

kbi s

TI
Rb=33.85x=
ls

bits
= 2080
fr"."

Figure: Mobile Station

J.60 ms

\ :-;::u.l /tms

= 1i.85 )r

^ sec
li bitst 160 samples

= s"mple

l-1.85

Speech Coding and lvlodulation Technique

Basic Wireless Conccpts

Frequency Concepts

The following table summarizes the frequency-related specifications of each of the

systenls. The terms used in the table are explained in the remainder ofthis section.
clated
Table-4: Freo uen
cations
P-GSM
E-GSM
9()O
GSM
I 8OO
GSM I9OO
Sj'stem
9OO
Frequencies

.
o

l..tplirk
Dori nlink
Bandrvidih
Duplex Disiance
Carrier Separation
Radio Channels
Transm ission Rate

890 9t5lvlHz
93i - 960 irti l.r
25 Mi{z
45 MHz
200 kHz

880

!,:i-

Ii

915
!160

MHz

t'ilir

I\{Hz

l7l0

l8C-l

1735 lvlHz

1850

iz

l()10

i3tiO

ir il

l9l0

60 lvlflz

95 lvlllz
200 kHz

80 lviHz
200 kHz

lli

45 MHz
200 kHz
r75

210 kbitsis

270 kbitsis

270 kbitsis

,15

NlHz

lq90 Nlllz

Illtz

75

GSN4

300

270 kbitVs

Prepired

by

Prol lltd Sarfur Rahnran

An lvlS communicates rvith a BTS by transmitting or receiving radio waves.

uhich consist of electromagnetic energy. The frequency ofa radio rvavc is the number of
times that the rvave oscillates per second. Frequency is measured in Hertz (Hz), where I l{z
lircclLrency:

indicates one oscillation per second, Radio frequencies are used for many applications in the
rvorld today. Some common uses include:
30C N{Ilz approx.
Television
FIll radio
- 100 l\{liz appiox.
Police Radio
- Country dependent
N'lobile Networks - 300 - 2000 MHz approx.

The frequencies used by mobile networks vary according to the standard being used. (As
these frequencies are used to carry information, they are often referred to as carrier
frequencies). An operator applies for the available frequencies or, as in the United States, the
operator bids for frequency bands at an auction. The following diagram displays the
frequencies used by the major mobile standards.

l"1fl? 0

450
NIIT.tso

800
AMPS

DA|VJPS

800

900
GSM 9OO
NMT 900

1800'19C0

1500
PDC

1500

GSN1

1800 oAMirs

1300

GSM 19C0

PDC BOO TACS

Figure : Frequencies for major mobile standards.

Note: Due to frequency, a BTS transmitting information at 1800 MHz with an output power
of l0 Watts (W) will covcr only half'thc arcn of n similar B'lS transnritting at 900 Nlllz. To
counteract this, BTSs using 1800 MHz may use a higher output power.
Bandwidth
Bandwidth is the term used to describe the amount of frequency range allocated to one
application. The bandrvidth given to an application depends on the amount of available
frecluency spectrum. The amount ofbandwidth available is an important factor in determining
the capacity ofa mobile system, i.e. the number ofcalls rvhich can be handled.
Channels
Another important factor in determining the capacity of a mobile systenr is the channel. A
channel is a frequency or sct of l'requencies lvhich can be allocated for thc transmission, and
possible reception of information. Communication channels of any form can be one of the

following tlpes:
Tlpe
Simplex
Half duplex
Full duplex

Description

Examples

One way only

Trvo way, only one at a time
Two way, two lines at a time

FM radio, television
Police radio

Mobile systems

A simplex channel, such as an FM radio music station, uses a single frequency in a single
direction only. A duplex channel, such as that used during a mobile call, uses frvo
frequencies: one to the MS and one from the MS. The direction from the Iv{S to the net\.\'ork
is refened to as uplink. The direction from the network to the MS is refened to as downlinh.

frcparcd

try IfoI

Nld SJIful IiJhn\'lrl

ru

a radio channel
Figure: Uplink and dorvnlink on

to transmit
Because it requires less power
frequencies in mobi'le tytt"*t
battery Porver of the MSs

"'"Ti*;;;

a lower frequency

trt"

i"**

ti,ra

"i

a siven distance' uplink

-over
ftequen-cies - this saves valuable

Duplex Distance

il.r *o'.1:,r::":lH'"'ll
in frequency bY a mtntmum. ot

rhe use of

;;;;ii;';;:q;;";ies

Hil\Jlf"d;;;]j:-,xH1:\;nHi.'iliK'i::
otn"

:.r, eacn
^^^L ^+L6,
^ with
would interfere

, - ---

915

935

Duplex Distance (45 MHz)

Bandwidth (2 x 25 MHz)
Figure: DuPlex Distance

Carrier SeParation

separation (Camer
mobile system includes a carrier
every
distance'
duplex
the
"to
'iunO*iAtftl This is the distance on the
ln addition to

separation

is

sometimes

t"tn"Jl

'it"""J"'

:l'*tT?J#'"::jffi il"jiil;i'*1;f;

lli:lln{ll*::*,:#:l;ffi

fr:Ji:'T.";iTn:*:lJ;'xff"T,:"trn?:Llt{?:*#:ll:'ff
#il m"" gr"ui"' ttt" amount of separation required'
MHz
Carrier Separation (200

895

895'6

KHz) t---'

Figure: Catrier SeParation

171

;requiredin

il:T;:'i;*mlt

Prcparcd br

Prof i\1d S.rrfrrr

Ruhnr.rn

From the Figure above, it can be seen that it can be scen that the infomration to be sent is
modulated around the carrier frequency of 895.4 MHz. The same is tme of the information to
be sent on 895.6 MHz. To avoid interference between the flvo sets of information, a
separation distance of 200 kHz is required, If less separation rvere used, they rvould interfere
and a callet on 895.4 MHz may experience crosstalk or noise fiom the caller on 895.6 MHz.

Capacity and Frequency Re-Usc

It is the number of frequencies in a cell which determines the cell's capacity. Each company
with a license to operate a mobile network is allocated a limited number of frequencies.
These are distnbuted throughout the cells in their network. Dcpending on the traffic load and
the availability of frequcncies, a cell may have one or more frequencies allocated to it.

It is important when allocating frequencies that interference is avoided. Interference can be

caused by a variety of factors. A common factor is the use of similar frequencies close to
each other. The higher the interference, the lower the call quality.
To cover an entire country, for example, frequencies must be re-used many times at different
geographical locations in order to provide a network with sufficient capacity. The same
frequencies cannot be re-used in neighboring cells, as they would interfere lvith each other so
special pattems of frequency usage are determined during the planning of the netrvork.

Figure Ncighboring cells cannot have the same frequency (simplified).

These frequency-reuse patterns ensure that any frequencies being re-used are located at a
sufficient distance apart to ensure that there is liftle interference betrveen them. The temt
"frequency re-use distance" is used to describe the distance between two identical frequencies
in a re-use pattern. The lower frequency re-use distance, the more capacity available in the
network.

KEY IDEAS OF MOBILE CELLULARSYSTEM

The basic principle of a cellular mobile radio is to split the requtred coverage area lnto a
number of smaller areas or cells, each of which has its own radio base station. The cells are
grouped together into clusters and the radio channels available are allocated to each cluster
according to a regular patlern, which rcpcats over thc wllole covcragc urca. ln this rvly, clch
channel is utilized several times throughout thc coveragc arca in a rcgular fashion. The
number of cells in a cluster has to be chosen so that the clusters fit together into contiguous
areas. Only certain cluster configuration can do this. A typical anangement of interest to
cellular radio is groups ol 4, 7, 12 and 21 cells, The number of cells in each cluster has
significant effect on the capacity of the overall system. The smaller the number of cell per

172

Prcparcd

by

Prol l\ld Sarliu Rahnrrrr

clustcr. th!- llrgcr thc numbct of channels per cell, thus thc traffic canied
Der cell is hiehcr.

Ho,,vever rcducing the cluster size increases the interference effect.

Bas

ic Cel ILtlar

sys tenr

A basic cellular system consists of three parts;

o A mobile unit: It contains control unit, a transceiver, and an anrenna.
r A ceil site: The celi site provides interface between the MTSo and the mobile Lrnits. lt
has a radio base station: control unit, radio cabinets, transmifting and receiving
antennas, a power plant and data terminals. The base station serves as a bridci
betr'veen all mobile users in a cell and connects the simurtaneous ce s via telcphoie

'

lines or microwave links to the MSC.

Mobile Telephone Srvitching office (MTSO): It is also krown as Mobile Switching
center (MSC), which is the central coordinating element for all cell sites. It contain!
cellular processors, cellular switch, subscriberi' information etc. lt interface with
telephone company offices, contrors calr processing and handles billing activities erc.

\t5(

Figurc : A cellular network.

Figure above shows the cellular systems with connections. Radio links and hich-speed data
links using coaxial cable or optical fiber connect different subsystems, say, BS t-o nasc, usc
to land telephone network etc.

Ccll Shdpe in Cellular Systent

The region of coverage of a base station is called a cell. The hexagonal shape sho*'n in
Figure below is a simplistic model of radio coverage for each base station. It might seem
natural to choose a circle to represent the coverage area of a base station. But adjacent circles
cannot be overlaid upon a map without leaving gap. Thus rvhen considering geometric shapes
which cover an entire region without overlap, the sensible choices are a square, an equilatiral
triangle and a hexagon. The hexagonal shape is used in practice because:
o For a given distance between the center of a polygon and its farthest points, hcxagon
has the largest area. So by using hexagonal geometry, the fewest number cells coirld
cover a geogmphic area.
o fhe hexagon closely approximates a circular radiaticn pattcm, rvhich rvould occur for
an omnidirectional base station antenna and propagation in free space.
Since actual radio coverage of a cell depends on many factors (such as geometric structure:
buildrngs, hills etc., traffic demand at differentparts etc.) and nor circulal hexagonal-shaped
tt)

I)r(.f.rr!J b) - l,fdl l\1J. Srilur tt.rhrnrn

cells arc ar-tificial and such a shape cannot be generated in the real rlorld. The real cell shapes
along with ideal ones are sholvn in the follolving figure.
|1e^aaOnAiCeliS

:'i

.\-

,'\
:'!ii
'\/

,.

lii

t.'l
'\-'^'-',t-

\+/
jl

'-."--.\
l

l.'-,'/\'-'./

---

Sigoal coverage

Figure:

Figure : Hexagonal cell.

Real and ideal cell shaPe.

Concept of Frequency Reuse

cellular radio systems rely on an intelligent allocation and reuse of channels throughout a
coverage area. Each cellular base station is allocated a group of radio channels lo be used
rvithini small geographic area called a cell.Base stations in adjacent cells are given channel

groups that contain completely different channels than neighboring cells. The base station
int.nnu. are specified to achieve the desired covefage within the particular cell. By limiting
the coverage area within the boundaries of a cell, the same group of channels may be used to
cover different cells that are separated from one another by distances large enough to keep
interference levels within tolerable limits. The design process of selccting and allocating
channel groups for all of the cellular base stations within a system is called frequencl'
planninglUsi of same frequency in a number ofcells lvithin the area to be covered is called

frequency reuse.

Figure : FrequencY reuse concePt

A particular radio channel, say F1, is used in one geographic zone called cell, say 6.1. with a
coverage radius R can be used in another cell with the same coverage radius at a distance.D
cells)
away. in this frequency reuse system, users in different geographic locations (different
of
concept
the
illustlates
Next
Figure
channel.
may simultaneously use the same frequency
group
channels.
of
the
same
use
freoLrcnct, l.",,r" *h"r. cells labeled rvith the same letter

111

TPrcpaIcd

by

Pfol: l\ld S.rliu lirhrn.ur

Figure : Cellular ffequency reuse concept.

Frequency Reuse Schentes Suppose, a cellular system has a total of S duplex channels
available for use. lf each cell is allocated a group of t channels (ft<$, and if the ,S channels
are divided among N cells into unique and disjoint channel groups, rvhich have the same
number of channels, the total number of available radio channels can be expressed as

S=lN

The N cells which collectively use the complete set of available frequencies is callcd a
cluster. lf a cluster is replicated M times within the system, the totnl number of duplex
channels, C, can be used as a measure of capacity and is given by

C=MkN=MS
As seen from the above equation, the capacity of a cellular system is directly proportional to
the number of times a clustcr is replicated in a fixed service area. The factor N is callcd the
cluster size and is typically equal to 4, 7 , 12, 19 etc.If the cluster size N is reduced rvhile the
cell size is kept constant, more clusters are required to cover a given area and hence a more
capacity (a larger value of Q is achieved. A large cluster size indicates that the latio between
the cell radius and the distance between co-channels D is small. Converscly, a snlall clustel
size indicates that co-channel cells are located much closer together, thereby incteases the cochannel interference. From a design viewpoint, the smallest possible value ofly' is desirable in
order to maximize capacity over a given coverage area (i e., to maximize Q The frequency
rertse factor is given by l/N , since each cell within a cluster may only lrse up to l/N of the
total available radio channels in the system.

ofthc figurc above has six cquidistant ncighbors

and that the lines joining the centers of any ccll and cach of its neighbors are separated by
multiples of 600, and there are only certain cluster sizes and cell layouts which are possible.
Due to the fact that thc hcxagonal geomctry

In order to connect the cells without gaps between adjacent cells - the geometry of hexagons
is such that the number of cells per cluster, N, can only has values, which satisfy the
follorving equation
N =i, +ij + j,
u,here / and j are non-negative integers. To find the nearest co-channel neighbors of a
particular cell, one must do the following:
I

/)

Prcpared by

- Prof Nld. Sarfur lirhntrn

(r)
(ii)

Nlove r cclls along any chain ol'hexagons and then

Tum 60 degrees counter clockwise and moveJr ce s. This is illustrated in the Figure
bclolv for i = 3 andj = 2 (N = t9).

Figure : Method of locating co-channel cells in a cellular system.

Frequency Reuse Distance
The minimum distance which allows the same frequency to be reused will depend on many
factors, such as, the number of co-channel cells in the viciniry of the center c;ll, the
[pe of
geographic contour, the antenna height, and the transmitted power at each cell site.
The frequency reuse distance can be determined bv

J:lrn

where N is the ftequency reuse pattem.

Theoretically, a large ly' is desired as the chance of co-channel interference is reduced.

Horvever, the total number of allocated channels is fixed. when ly' is too large, the number of
channels assigned to each of N cell becomes small, which results in decrease in trunkins
efficiency. Norv the challenge is to obtain the smallest number N, which canmeet the syste;
performance requirements. This involves estimating co-channel interference and selecting the
minimum frequency reuse distance D to reduce co-channel interference.

t16

Preparcd

by

Prof

Ild

Srrfur Rahman

Figure : Frequency reuse distance.

Example B,l: If a total of 33 MHz of bandwidth is allocatec! ro a particular FDD cellular
telephone svstenr y,hich uses t'wo 25 kHz simplex channels to provide full cluplex t,oice and
cor rol clnnnels, conpute the nunber ofchannels available per cell tfa st,stem uses (a)fourcell reuse, (b) seven cell reuse, and (c) I2-cell reuse. If I lvlH; of the allocatet! spectrun is
cledicatecl lo conlrol channals, determine an cquitable clistribution of co trol chautels qnd
voice channels in euch ccll ,for eaclt of tlree n'slclas.

Channel Assignment Strategies

For efficient utilization of the radio spectrum, a frequency reuse scheme that is consistent
with the objectives of increasing capacity and rninimizing interference is required. A variety
of channel assignment strategies have been developed to achieve these objectiVes. Two t)?es
of strategies are
. Fixed channel assignment and
. Dynamic channei assignment

Jixcd chunnel assignntc t stt'(ucg)', each cell is allocated a predefined set of yoice
channels. Any call attempt rvithin the cell cln only be sen'ed by unused channels in that
particular cell. If the all the channels in that cell are occupied, the call is blocked and the
subscriber does not receive any service. Several variations of the fixed assignment strategy
exist. In one approach, called borror.ving strategy, a cell is allowed to borrorv channels from a
neighboring cell if all its orvn channels are already occupied. The N{SC supewises such
borrorving procedures and ensures that the borrowing of a channel does not disrupt or
interfere rvith any of the calls in progress in the donor cell.

In

the

In a dl,namic channel assignment strategy, voice channels are not allocated to different cells
pemlanently. Instead, each time a call request is made, the serving base station requests a
channel from the MSC. The l'4SC then allocates a channel to the requested cell following an
algorithm that takes into account likelihood of future blocking within the cell, the frequency
of use of the candidate channel, the reuse distance, and other cost fr:nctions. Accordingly, the
MSC only allocates a given frequency if it is not in use in the cell or any other cells, rvhich
Ltl

f'rcparccl

by

Prot i\ld. Surlirr liahntrn

restricted distance of frequency reuse to avoid co-channel

interference. Dynamic channel assignment reduces the probability of blocking, rvhich
increases the trunking capacity of capacity of the system. This strategy requires the NISC to
collcct the real time data on channel occupancyi traffic distribution, and radio signal strength
indication /RSS1) of all channels on a continuous basis. This increases the storase and
cornputational ioad on the system but provides advantage of increased channel utilization and
decreased probability of a blocked call.

fall within the minimum

,.,*"l
,.r.t.

vi
il

Problem 1: If a total of 5MHz bandwidth is allocated to a cellular system (FDMA), which

uses 25 kHz wide channels for both voice and control, compute the number of channels per
cell if the system uses 7-cell reuse. Also, calculate the total capacity if fiequency is reused 20
2oo x2o = h olt
trmes
Problem-2: Suppose that a mobile operator has been allocated a 5MHz bandrvidth by the
BTRC. Compute the total number of channels for voice antl control in a cluster. If the cluster
has 7 cells, suggest a realistic channel distribution in each cell. Also, calculate the total
capacity of the ststem ilthe frequency is re-usea zO iimei.'Alsume GSM standards' rrhere

,L the carrier separation is 200kHz and 8 time slots are used per camer.

99*?O= Eoo

eotiS;99

Transmission Rate

In GSM 900, the frequency that is used to transfer the information over the air interface is
around 900 MHz. Since this is not the frequency at which the infomration is generated Since
this is not the frequency at rvhich the information is gencrated, nodulation tcchniques arc
used to translate the information into the usable frequency band. Frequency translation is
implemented by modulating the carrier wave in accordance with the lvavefoml of thc input
signal (e.g. speech).

The modulation technique used in GSM is Gaussian Minimum Shift Keying (GMSK).
GMSK enables the transmission of 270 kbits/s within a 200 kHz channel.
Access Method: Time Division Multiple Access

(TDMA)'

Most digital cellular systems use the technique of Time Division Multiple Access (TDMA) to
transmit and receive speech signals. With TDMA, one canier is used to cany a nunrbcr of
cells, each cell using that carrier at designated periods in time. These periods of time are
referred to as time slots. Each MS on a ca'll is assigned one time s'lot on the uplink frequency
and one on the downlink frequency. The information sent during one time slot is called a
burst,

In GSM. a TDMA frame consists of 8 time slots. This means that a GSM radio carrier can
carrv 8 calls.

I'rcparcd by

Prof. i\ld. Sril'ur R{hmtrn

Figure: TDMA

Note: Only the downlink direction is shown. There is also a corresponding

frame in the
uplink direction.

Group Speclal Mobile (GSM) : RADIO ASPECTS

Multiple access technique

Uplink frequency
Downlink Frequency
Channel Bw
No. of channels
Channels / canier
Frame Duration
Interleaving Duration
Modulation
Snecnh.^.1i..
Speech coder

: TDMA

/ FDMA / FDD

fu
fd
B

= 890 -91 5 MHz

= 835 - 860 MHz

N
N"

= t74
= 8 (Full Rate) , l6 (Half Rate)

=200 KHz

Tf

= 4.6 ms

= 40ms

GSMK 4149K
RPE - LTE convolutional
13 kb/s (full rate)

Bit Rate

Associated control ch

Extra

Hand of scheme

Mobile assorted
0.8 w, 2.5 w, 8w

MS power level

960

.)

Frame

(Dnlink)
935
I

4.6 ms = T1

t79

0.5 75

l)repafcd

Ilundu.idth

: [J/l_ : 9li -690

D/L : 960 -935

BW ofeach channel = 200 KHz

25

MHz

b)

I)rot. iVkl Srriirr luhman

=25 MIlz

:25 Ml.lz

25000

- f00 - 124 chrnnel

200 NlH,
Each uscr transmits in each slot for a dLrration of
No. ol

chrnnel

-lr=STr 4.6
g rs=0.575 ms

full Rate Tx = 8 Timc Slots => 22.8 kbps per slot

Half Rate Tx = 16 Time Slots=> l l.4 kbps per slot
Basic Traffic Cases

Traffic Cases: r\iS in Idle Mode

2.
3,
4.
5.
6.
7.
8.

Ftgure: Traffic Cases rvhen MS is in idle mod..

IMSI Attach
Location Updating: Tlpe IMSI Attach
Changing Cells within an LA
Location Updating: Within the same MSC/VLR
l.ocation Updating: With a ncw MSC,^/LR
Location Updating: Type Periodic Registration
IMSI detach
Implicit detach

Attaching to the Netryork

lntcrnational Nlobile Subscriber Idcntity (IMSI) Attach
Every subscriber in a GSM netrvork is identified by a unique number called the Intemational
lv{obr}e Subscriber Identity (IMSI). The IMSI is used rvhenever an N{S is in communication
rvith

Prepared by

Prof. i\ld. Sarfur Rahman

'when an MS is s"vitched on. the IMSI attach procedure is

executed. This invorves
following steps:

t
2

il:"YT:las

the

an IMSI auach message to rhs nctwork indicaring

rhar ir has changcd

The vLR determines whether there is a record

for the subscriber arready present. If
contacts the subscriber,s IILR for a copy
of the subscription

il"r[::,#-

3. The VLR updares the MS status to idle.

4. Acknowledgement is sent to the MS.

,r/,
.//

" /,/V

Figure: IMSI Attach

Location Updating: Typ IMSI Attach

If

the MS has changed LA while powered ofrr the IMSI

attach procedure may lead to an

update to the location of the MS.

During IMSI attach, the vLR may determine that the

cunent LAI of the MS is different from
the LAI stored in the MS's subsiription information.

If

MS,

;;;ih;

vLR

updates the

LAI of the

Roaming in the Network

Changing Cells lVithin an LA
MSs are constantry moving around in the cellurar
network. r'he MS location informarion
stored in the vLR is the LA. If an MS changes
cells within an LA, the netrvork is not
updated.

The MS knows that the new cel-belongs to the same

LA by listenrng to the contror channel
frequency in the new cel. This broadcasts the cell,s
LAI.'T;; MS compares the rast LAI
received with the new LAI. If they are the same, it
n-,run, tt ut tf," MS has not changed Las
and does not need to inform the network.

Location Updating: Within the same MSC/VLR

MS detects a change in LAI on the control channer, it informs
the network. when the
updating message, the MSC/vLR determines whether it is
-Location
an IvfS
rvhich is already
registered or on MS visiting from unott., nliC,lrfR.

Jlln

MS

sends. the

l8l

l,rcprrcJ b)

- l'rol irltl

SJrlirr llJhnlrrr

Figure: Location Updating: within the same MSC/vLR

l.TheMSlistenstothecontrolchannelinthenewcelltodeterminetheLAl.Thenerv
LAI is compared to the old one lf they differ, a location update is necessary'
2. The MS establishes a connection with the network and authentication of the
subscriber and MS is Performed'

3. lf

Request to the
authentication ts successful, the MS sends a Location Upfdating

system.

,1. The system acknou4edges Location Updating

Location Updating: \Yith a Nerv MSC/'VLR
is pcrfomred llorvcvcr' unkno\\'n to
When an MS roams Into a ncw LA, location updating
Request is
belong to a new MSClr'LR When the Location Update
the MS, the LA may
below'
received by the nerv VLR, it executes the procedure

t ,"/

s
Figure: Location Updating: With a new lvlSC/VLR

1.

2.

:.
4.

successful' the VLR checks its

Authentication is peribrmed lf authentication is
for this MS'
dutobur" to d"t"r.in" whether or not it has a record
a request to the subscriber's [lLR
When the VLR finds no relord for the MS, it sends
for a copy of the MS's subscription'
its location information for
rir. nr-i passes the i"io"rot-i'". to the vLR and updatesthe
information it has on the
to delete
the subscriber. The HLR inst*ctt the old VLR
MS.
for the MS' including the latest location
The VLR stores its subscription information
to the MS
anJ status (idle). The VLR iends acknowledgement

Location Updating: Type Periodic Registration

NISs to send a reqistration message to thc
Periodic registration is a f'eatr-rre rvhich forces
miss such a regrstration' the netu'ork rvill
netlvork at predefined intewals lf an MS should
out of the area of coverage and ensures
,lr".f.-in" M.S as detachecl This may occur if an MS is
thrt ncedlcss paging is not performed

f'r'cpared by

- i'rof \ld

Sartut Rrhnran

If the nenvork uses periodic registration, the MS will be informed, on the control channel of
lrorv oftcn periodic registration must be performed. Periodic registration has an
rcknorvlcdgcment mcssilgc. The MS trics to register until it receives this message.

Detaching From The Nctwork

In-ISI Detach

IlvlSl detach enabies the MS to indicate to the netrvork that it is switched oft'. At porver off,
the MS sends an IMSI detach message to the nefwork. On reception, the VLR nrarks the
corresponding IMSI as detached. The HLR is not informed. No acknowledgement is sent
to the NIS.

Implicit Detach
If the MS sends an IMSI detach message to the System and the radio Link quality is poor, the
system nlight not be able to decode the information. Because no acknowledgement is sent to
the MS, no further attempt is made. In this case, the system stili regards the MS as anached.
lf periodic registration is in usc, the systcm rvill soon dctctlrinc llat thc NIS is dctlchcd. 'l hc
VLII then pcrforms an inrplicit dctach, rnarking tlrc MS as dctachcd.

Traffic

Cases:

I{S in Active }Iode

An MS is in activc mode rvhen there is a call (speech, fax or data). or a call setup procedure
takins Dlace.

Figure: Cascs rvhich activate an N'lS and cascs when NIS is in active mode.

Call liom an NIS to a I'STN Subscriber

'fhis scction describcs wltat happens when a mobile subscriber wants to set up a voice call to
a sLrbscriber in the l'S'fN.

183

Prcpared by

- Prof NId Sarfur Rahman

1.ra

.;

\s,i

,/

,4/

ts
Figure: Call set-up MS to PSTN

I . The MS sends a call set-up request to the MSC/VLR. The MS

and MSC/VLR then
communicate to set up the call:

.
.
o

Marking the MS as "active" in the VLR

The authentication procedurc
Sending the B-subscriber's number to the network

2. The MSC/VLR instructs the BSC/TRC to allocate idle time siot on a TDMA frame in the
cell. The RBS and MS are told to tune this time slot.
3. '1 hc MSc/Vl-R {brwartls thc B-nunrber ro an exchange in rhc Fs't'N, which catobtishca
r
connectlon to the subscriber.
4.

If

the B-subscriber answers, the connection is established.

Call from a PSTN Subscriber to an MS

The major difference between a call to an MS and a call from an lvIS is that in a call to an lvls
the exact location of the mobile subscriber is unknown. Therefore, the MS must be located
using paging before a connection can be established.

Belorv is the description of the call sefup procedure for a call from a pSTN subscriber to a
mobile subscriber. A call from an MS to a mobile subscriber operates according to the same
process, the only difference being that the GMSC is contacted by another MSC/VLR instead
ofby a PSTN node.

/7
T
fl
ti

fl.1

8o

'\.
\ \\ {[

igure:: Call to MS frorn PSTN

I 84

Pfcpared by

Prot: Md. Sirifur Rahntin

l. The PSTN subscriber keys in the MS's

telephone number, rvhich is analyzecl in the

PSTN which identifies that this is a call to a mobile nehvork subscriber. A connection
is established to the MS's home GMSC.
2. The GMSC analyzes the dialed number to find out which FILR the MS is reeisrered
in, and queries the HLR for information about how to route the call to rheiervinc
MSC,^/LR.
The HLR retunts information about how to route the call to the serving MSC/VLR.
The GMSC ror.rtes the call to the MSC/VLR.
4. The MSC/VLR knows which LA the MS is located in. A paging message rs senr ro

5.

the BSCs controllins the LA.

The BSCs distribute the paging message to the RBSs in the desired LA. The SRBSs
transmit the message over the air.
6. When the MS detects the paging message, it responds and call sefup procedures are
executed. A time slot is allocated and tuned to.
7. The mobile phone rings. If the subscriber answers, the connection is eslablished.
Handover
The process of changing cells during a call is called handover. To choose the best tareet cell.
measurements are performed by the MS and the RBS.
Locatirrg

An MS continuously measures signal strength and quality on its own cell and siglal strength
on the control channel carries of the neighboring cells. The measurements are canietl on the
downlink while MS is in active mode. The measurement results are sent to the BSC at recular
intervals. The scrving RBS rneasures signal strcngth and quality on the uplink
RBS measums:
Signalslrngu and
transmlsslon qualitY
onTCH. uDllnk

Mcasuremants lrom RBS End lvls

S*
MS measurcst

Slgnalslronglh and

Signal str.nglh

transmisslon quaritY

onTCH.downlink

8TS

Figure: Measurements sent to BSC

Based on these measurements, the BSC decides if a handover is necessary and to which cell.
This is called locating. As soon as a neighboring cell is considered to be bener than the

serving cell, a handover is attempted.

Thcre are several types of handover, including:

o
o
r

Handover betrveen cells controlled by the same BSC.

Handover betrveen cells controlled by different BSCs, but the same MSC/VLR
Handover between ce'lls controlled by different MSC/VLRs.
185

l'rcparcd by

Prof. Nld. Sarfur Rahmln

Handover bet*'een Cells Controlled by the Same BSC

\\rhen performing a handover between two cells controlled by the same BSC, the MSC/VLR
is not involved. However, the MSCIVLR will be informed when a handover has taken place.

If

the handover involves different

I,4as,

iocation updating is perfonned once the call is

finished.

"b_=
Figure: Handover: Cells controlled by the same BSC.

1.

2.

3.
4.

The BSC orders the new RBS to activate time slot.

The BSC sends a message to the MS, via the old RBS, containing information about
the fiequency and time slot to change to and also the output power to use.
The MS tunes 10 the new frequency and conect time slot. The MS sends a thndover
Complete message to the BSC via the new RBS.
The BSC tells the old RBS to release the old time slot.

Handover between Cells Controlled by Different BSCs but the Same NISC/VLR
When another BSC is involved in a handover, the MSC/VLR must also be involved to
establish the connection between the two BSCs. Otherwise, the handover process is identical
to the nrevious handover traffic case.

FigLrre: Handover: Differenl BSCs but the same MSC,/VLR.

Preparcd bv

Prof.

i\kl S: firr Rlhmrn

lllndover betlvccn Cclls Controlled by Differcnt MSC/VLRs

be performed n'ithtn
Handover between Cells Controlled by Different MSC/VLRs can only
N4SCTVLRs also
different
by
one PLMN and not between hvo PLMNs. Cells controlled
means that they are controlled by different BSCs'

Figure: Handover: Cells controlled by differcnl N{S(ls'

TheoldMSC,MSC-A,retainsmaincontrolofthecalluntilt|recalllscleartd.Thisis
b".uu,.itcontainstheinformationaboutthesubscriberandcalldetailsSuchaScharging'
never belongs to
After call release, the MS mr'rst perform location updating because an LA
by the VLR-B' and rvill in tum
more than one MSC/VLR service area The HLR is updated
tcll VLR A io delete all infornation about the mobile subscriber'

Intcrnational Traffic Cases

oneoftheprimaryfeaturcsofGsMisthcabilitytoperfonlllrrtcmationalroantittgandto
rvhilc
to bc ablc to makc calls
handle lntemational call cases. ln ordcr for a mobiic subscribcr
subscriber's
io".ing in a different GSM network, there must be an agreement between the
operator'
home network operator and the visited nenvork

cases' the trvo traf'fic cases *'hich

Although there are less significant affects on other traffic
are affected most are or.ttlined belolv
IMSI Attach

When an MS is roaming Intemationally, the follorving occltrs:

within one frequency band

The MS is switched on and scans all GSM frequencies
a control channel carrier' The MS tunes to the
ie.g. CSU 900) It is searchingfor
s)'stent
carrier which ias the stronger signal strength and reads its
"oitrol
"honn.i
information. This inclLrdes the identity of the network operator'
in the SIM
2. The MS compares the network identlty with the list of forbidden PLMNS
home
subscriber's
memory. This list contains all network identities which the
rvhich
with lf the network
operator does not have an Intemational roaming agreement
to scan for a permitted
continues
MS
thi
network'
thc MS has tuned to ls a forbidden

1.
"

netrvork.

187

PreparcJ by

Prof.

Ild

Sarfur Rahnran

l. Il thr": MS docs not fincr a permittc<J nctwork, but rras idcntified a forbiddcn nctwork, il
4.

displays the nressage "Enrergency ca s only".

tuncs to it and scnds an IMSI attach messase.

If thc MS finds

a pemrittecr

n"n"o.i, ii

Thc remaindcr of rhis traffic clse is i,icntrcll to thilt of thc rron.al INISI d[]ch casc,

u.'ith thc only differcnce being that thc subscribcr's FILR is locatcd in
anothcr corntry.

Call to an MS

ar N{S is roaming Intemationalry and a ca is made to it, the procedure used is

identical to when the MS is in their home nehvork. The only major difierence
is that the
GMSC and HLR used are in the home netw,ork, while the tr4SC,lfln is in a network
in
another country.
)l'he1

Transmission Problems
\'lany' problcms may occur during the transmission of a radio signal. Sor.c
of thc nrost
conmon problems are describcd bclorv.
P

ath Loss

Path loss occurs r.vhen the received signal beconres r,veaker and rveaker due
to increasing
drstance behveen the I\4s and RTS, even ifthere are no obstacres bet\veen the
transmitter (Txl
and.ccei'ing (Rr) antcnna. The path loss problem serdom leads to a dropped car because
belbrc the problem becomes exlreme, a nerv transnrission path is estabrished via another
BTS,

l.

Shadowing

Shador.ving occurs rvhen there are physical obstacles including hiils and buildings betrveen
the BTS and the MS. The obstacles create a shadowing efibct rvhich can decrease the
recei'ed signal strength. when the MS moves, the signal strength fluctuates depending on the

obstacles between the MS and BTS.

-"\ signal influenced

dips.

by fading varies in signal strength. Drops in strength are called fading

Figurc: Shadorving

188

Prcparcd by,- Prof

i\jd

Sir

r Rrhman

2. iVlulti-path Fading

Multi-path fading occurs when there is more than one transmissron

paths to the rvls or BTS,
and therefore more than one signal arriving at the receiver.
This may be due to buildings or
mountains, either ciose to or far from the reieiving device.
Ral,lcigh

fading

and

tine dispersiarr are forms of multi_path fadrng.

Rayleigh Fading

This occurs when a signal takes more than one path between
the MS and BTS antennas. In
is not received on a line or rigrtt puth iirectty fronr the Tx
a.tennd.
Ilather, it is reflected off buirdings, for exampre, ina'is .ec.ru.i
rron., ."u.rot Jrii"r.ii
indirect paths. Rayleigh rading occurs when the obsiacres
are crose to the receivins

this. case, the signal

antenna.

Figurc : Raylcigh Fadrng

Time Dispersion

Timc drspcrsion is another problem rerating to murtiple paths to the

R\ antenna of either an
MS or BTS. Horvever, in contrast to Rayleigh fading, ihe reflected
signal comes from an
object far away Ilom the Rx antenna.

iliH

A\

"' -!*

Tr.

A''

A
m

-q
\

Figure: Time dispersion

189

_!\

Plcprrcd b1 - Prof Nld Sarfur Ilihman

If the reflected signal arrives one bit time after the direct signal, then the receiver detects a I
from the rcflected wave at the same time it detects a 0 from the direct rlave Thc slmbol I
interferes rvith the symbol 0 and the MS does not knorv lvhich one is corrcct
3. Time

Alignmcnt

Each MS on a call is allocated a time slot on a TDMA fiame/ This is an amount of time
during which the MS transmits information to the BTS. The information must also arrive rt

rhe BTS rvithin tirat time slot. The time alignment problem occurs when parl of the
informatiorr transmitted by an MS does not arrive within the allocated time slot. Instead, that
part may arrive during the next time slot, and may interfere with information from another
lvlS using that other time slot.

,b

Time alignnent is caused by a large distance befween the MS and the BTS, Effectively, the
signal cannot travel over the large distance within thc given time

TDMA-framc

TS0

i rsl I

Ts2

B, Far away
from BTS

NN

B.WZ

A
E'IJ

A, Close to BTS

IrigLrrc:'fhc timc alignmcnl pfoblcnl

For example, an MS is close to a BTS and has been allocated time slot 3 (TS 3). During the
call, the MS moves away from the BTS causing the information sent flom the BTS to arrive
ar the lvIS later and later. The answet ftom the MS also arrives later at the BTS. If nothing is
done, the delay becomes so Iong that the transmission from the MS in time slot 3 overlaps
rr ith the information, which the BTS receives in time slot 4.

Combined Signal Loss

Each of the problems described above occut independently of each other. However, in most
calls some oi these problcms may occur at thc sanc timc. An illustration of \Yhat the stlength
may look like at the MS R{ antenna when moving away from the BTS Tx antenna is shorvn
in iigur. below. The problems of path loss, shadowing and Rayleigh fading are present for
the transmission path.

The signal strength as a global mean value decreases with the distance (path loss) and finally
results in a lost connection. Around this global mean, slow variations are plesent due to
shaclorvinq effects and fast variations are present due to Rayleigh fading

19t)

Prcpxrcd by

- Prot \td Srrfur

R.rhnran

=
.9

log (dislance)

Figure : Receiver (Rx) Signal strength versus disrance

Solutions to Transmission problems

This section describes some sorutions to the problems described in various
sections. Although
many of these do not entirely solve all problems on the radio transmission path,
thcy do plilan rmportant part in maintaining call quality for as long as possible.

Channel Coding

All

signals, analog and digital, become distorted over distances. Because of

some unique
digital transmission is preferred to its anarog counterpart. In analog, the only
solulion io_ this is to anrplily the signal. Flowever, in doingio, the diitortion
is arso"amplifiei.
In digital the signal can be completely regenerated as new, without the distortion.
advantages,

Figure: Regeneration of digital signal

ln digital transmission, the quality ofthe transmitted signal is often expressed in terms
ofhorv
many of the received bits are incorrect. This is called Bit Error Rate (BER). BER defines
the
percentage ofthe total number ofreceived bits which are incorrectly detected.

bits
bits
Errors

Transmitted
Received

I I 01000I 10
I 00I 00I0I0
1
11

3/10 = 30% BER

This percentage should be as low as possible, It is not possible to reduce the percentage to
zero because the transmission path is constantly changing. This means that thcre must be an

for a certain amount of errors and at the same time an ability to restore the
information, or at least detect errors so the incorrect information bits are not intemreted as
conect. This is especially important during transmission ofdatal rs opposed to speech, for
rvhich a higher BER is acceptable,
allorvance

channel coding is used to detect and co[ect errors in a received bit stream, It adds bits to
a
message. These bits enable a channel decoder to determine whether the message
has faulty
bits, and to potentillly couect the faulty bits.
191

I'rcpa|cd

by l'fof i\l(l

Srrlirr R.rl\rrlrn

Interleaving
In realit-v. bit errors often occur in sequence, as caused by long fading dips affecting several
consecutjle bits. Channel coding is most effective in detecting and correcting single errors
and short error sequences. It is not suitable for handling longer sequences ofbit errors.
For this reason, a process called interleaving is used o separate consecutive bits of a message
so rhrt these rre tr]nsmitted in a non-consccrrlire ury.
For example, a message block may consist of four bits (1234). If four message blocks must
be transmitted, and is lost in transmission, without interleaving there is a 25% BER overall,
but a 100% BER for that lost message block. It is not possible to recover an)'thing from the
lost block.
1

3 4

message blocks

iltedeaving

x$l

all 3 3

591

3 J

Inlerleaved
message blocks

Figurc: Intcrlcirvlng

lnterleaving could be compared to sending a group of important people from A to B on

different planes. By doing so, the likelihood oflosing the entire group is ntinirnized in case of
a plane crash.

If

interleaving is used, as shown in the above figure, thc bits of each block may be sent in a
non-consecutive manner. If one block is lost in transmission, again there is a 25% BER
overall. However, this time 25% is spread over the entire set of message blocks, giving a 259i'
BER for each. This is more rnanageable and there is a greater possibility that thc crrors cln be
conected by a channel decoder.
1

3 4

X 3 4

Figure: Received interleaved message blocks.

.\daptive Equalization
Adaptive equalization is a solution specifically designed to counteract the problem of time
dispersion. it works as follows:

L
2.

l.

as training seqttcnces. Thcse are

(programmed
at
the
time
of manufacturing). The BTS
known to thc BTS and the MS
instructs the MS to include one of these in its transmissions to the BTS
The MS includes the training sequence (shown in figure as an S) in its transmission to

A sct of predcfined known bit pattcms cxist, known

the BTS. Horvever, due to the problems over the radio path, some bits may

be

distorlcd.

Thc BTS rcceives the transmission from the MS and examines the training sequence
u,rthin it. The BTS conpares the received training sequence rvith the training
sequence rvhich it had instructed the MS to use. If there are differences betleen the
192

Prepared by

4.

- Prof i\ld 53riur Rrhnran

two, ir' can be assumed that the problems in the radio path affected these bits must
have had a similar affect on the non-training sequence biis.
The BTS begins a process in which it uses its knowledge of what happened
to the
training sequence to corect the other bits of the hansmission.

Figurc: Adaptivc cqualization

Because some assumptions are made about the radio path, adaptive cquarization
may not
result in a l00o/o perfect solution every time. However, "good enough" result rvill
be achie'ed

Frequency Hopping

It is possible for the BTS and MS to hop from frequency to frequency during a call. This
may
a void some Rayleigh fading.

rOMA

lr.rc F

Frgure: Frequency Hopprng

Timing Advance
is a solution specifically designed to counteract the problem of time
alignment. It works by instructing the misaligned MS to transmit its burst earlier than it
normally would.

Timing advance

*ii.,

F-'|il1E;Ll
T

t
f ;i; l;i{lJT;f,:l
's.
\$
I

ff

A
Figure: Timing Advance

193

Prcpdrcd b) '- I'ro[. Nld. Sililirr Rihmrn

Appendix:
\
joql

l. Two Rav Model for Mobile Radio Pronasation

Let.

9=?, (fu\

h' = Height of transmitting antenna

h, = Height of receiving antenna

\a

0 = Elevation of reflected ray

$ = Elevation

ofdirect raY

Received power, Pr = pt

2
t-fl
., o, 1t+a"^ jlf

l2

4\

d = distance between the tx and receiver

a, =

reflection coefficient of the ground reflected ray

= phase difference between the direct ray and ground reflected ray.

1=

wavelength of carrier.

I
h:

_t_
I
I

_L

hr

f
h,

L
D

1*A.e

Prepared b]"

Prol'. lvld. Saifur Rahman

Point-To-Point Tx Between Fixed

_L
h,

Tl.
n2

Received Power
Pr = Pt

( 1^
lr a".j0, -jA0 l,
d )' I
4\

Pt = Transmitted power

f=
)\

distance between two stations

= wavelength

An 0" = amplitudc and phasc of

a complex reflcction

cocfficicnt.

A0 = Phase difference caused by path difference Ad berween direct wave and the

reflected wave

=p^d=t^d
Complex reflection coefficient can be represented as

;n

"

V^l _=;.e. sin 0r- (e.- cos2 Or)ln

srn

ilt'"-;*'qiD

Wt,

(A)

V9:

Where e. is a dielectric constant that is different for different media

When vertical incidence is small

,0

is very small and

Dl

Pr

=--t - I l- cos Ag -j sin A{ |

(4n1

5= -l , {, = 0

),

= Fo (2-2 cos A{)

With effective antenna height of fr{ = nr +Hr and h{ = h: + H:, Ad can be found as
195

Prepared by

^d

Prof. Md. Saifur Rahman

=\(mr;4F

Since d>> h( r,

d>t

Ad-

n
u

,1.(
Ilrhz
----;'>-

h/z

.,./ .1.,
z /
(h'r+ h'z)'
(hi,,*
,
- _' h';)t
tr-r

217-

,t- r../ ./
rr

andAd=+Ad-r/rrr-r
'^d,\.

,o-*

'

Following five conditions result from the relation

Pr = Ps (2-2 cos A g)

Condition-l

Condition-2

: Pr =0 Then (2-2cosn$) =0or, AS<|

Condition-3

: Pr=PoThen (2-2cosAg) = I or, AO=1600=tl

Pr < po Then

lg<f

(2-2cosA{) <lor,

Condition-l: Pr>PoThen (2-2cosA{) > I or.

I .AA.}

Condition-5: Pr=4PoThen (2-2cosA$) =max or, AO=,1r

3. Land

- to - Mobile Transmission

over water

_F

;l

ul

h2

H2

_]There are always two equal strength reflected waves one from
the water surface and
one from the proximity ofthe mobile unit in addition to the direct
wave .
The total received power

196

Prepared by

Pr

I l-

(4)'

aQr and 402 are rhe phase

diffe

Dl

tar{ 1,

cos A$2

(AOr):

AOr , sin

cos (A0r)
sin

r'-

Then

Sarfur Rahnrar

-"ja0z l,

+jAOr

two reflected rays. As

ag1 ,

- prof Md

difference between
direct rav and the

^J:::nffath

cos aor - cos A0:

-

i(

sin aor + sinaoz)l

(Ag2):6g,

Pi

"=ll;

= po

-l

-j

( Aor

[t+ ( A0r + A0z)2j

In most practical cases

Then Pr

+^oz)i'z

-4-

(44),

AQ1

+ A0: < I

or

(A0r + A0:)2 <<

The equation is no1 same,

.rhe

that for power received

from the free space condition.
path-loss in u tuna-to.a
'mobile propagation
over water is 20 dB/dec.

For free space condition

n.=roftb;,
Which provides a path
loss of40 dB/dec.

Problem

l:

(a) What is the rec

Eansmitter power is
l w and ou)elved

of lmire(r.6km)

Power (in dBm) is the fiee space

ofa signal whose
is 2 4 GHz if the
receiver

oo, ,r. n"rr,*T-:lfieq'rency

tttt"t t Assume

as at a distance

fansmifte and r(
1.6 dB. (b) what is path
,o.. ,n dB ? (c) what i, tr,"
t *..i.,;on:11il:".T"
2 : What is the coverage

t"'^ ""

ofa base station that hansmih

u tt*u, at 2kw given
that the receiver sensitivity _100
is
dBm, the path loss at
th. o.r,' meter ls 32 dB
and path
ioss gradient is g =
4

froblem

197

Prcparccl by

Problem 3: Ifa base station covers

what would he coverage

if it were

- Prof

lr4rl.

Siltfur R.rhnrln

km in a plain area modeled as a two_ray channel.

used with a satellite ?

Appendix B
Why 800-MHz

spectrum is allotted?

The FCC chose 800 MHz because ofsevere spech'um limitations

at lower frequency bands.
FM radio broadcasting services operate in the vicinity of 100
MHz.
The TV broadcasting service starts at4l MHz and extends up
to 960 MHz.
Air-to- ground systems use l lg to 136 MHz; military aircrafts use 225
to 400 MHz.
The maritime mobile services are located in the vicinity of
160 MHz.
Fixed station services are allocated portions ofthe 30, io 100MHz band.
Therefore, it was hard for the FCC to allocate a spectrum
in the lower portions ofthe
30- to 400-MHz band since the services of this band become
so crowded. on the other
hand, mobile terephone transmission cannot be applied at 10
GHz or above because
Severe propagation path loss (rain drops, oxygen and other gases
or particles absorb
the energy)

.
r
.'
.
o

o
.

8.00

Multipath fading
Rain, storm etc.

was-originally assigned to educational rV

Ylz
th: load

:B:ed

channels. Forrunately cable TV service

ofproviding these TV channels. Then FCC allocated

MHz to the mobile radio system.

198

40-Muz system at g00

Prep;rrcd b1

- Prof Nld

Slrfur Rahmln

CDN,IA TECHNIQUES IN CELLULAR NIOBILE CONINIUNICATIONS

Code Division Nlultiple Access (CDMA)
-l-hc

third ty,pe of multiple access scheme is code division multiple access, rvhere each user
employs a parricular code address to spreacl the carrier bendb'idth over a much larger
bandrvidth so that the earth station community can transmit simultaneously without fre,rue;cv
or timc separation and with lorv interference. Before going to describc the operation of a
GDMA system into greater detail, we should first discuss about the spreacl sDcctrunl
technique.

Spread Spectrum Modulation Techniques

,'\ll of thc modulation techniques described so far strive to achieve greater po\\.er and,/or
bandrvidth efficiency in a stationarv additi\,e rvhite Gaussian noise channel. Since bandrvidth
is a linrited resource, one of the primary ciesign objectives of all the modulation schemes
detailed thus far is kr minimizc the required transmission bandrvidth. Spread Snectrum
techniques, on the other hand, employ a transmission ban<irvidth thrt is si'r'cral or,jcrs of
magnitude greater than thc nrininrum rcquired signal bandrvidth. \\rhilc this systcnl is \.r),
bandwidth inefficient for a singlc uscr, the advantagc of spread spcctrum is that man1. uscrs
can simultancously usc thc sanrc bandwidth rvithout significrntly interfcring ,r'iih one
another. In a mu)tiple-user, multiple access interlerence (MAl) en'rronmenr, spreao sp!'crnrnl

systems become very bandrvidth efficient.

Apart from occupying a very large bandu,idth, spread spectrum signals are pseudorandom
and han,e noise-like properties rvhen cornpared rvith the digital information data. The
sprcailing
is controlled by a pseudo-noise (PN) sequence or pseudo-noise code,
rvhich is a''avcfonn
binary scquence that appears random but can be reproduced in a detcrministic
manner by intended receivers. Spread spectrum signals are clcmodr.rlatcd at tlle receiver
through cross-correlation rvith a locally-gencrated version of the pseudorandom carrier.
cross-correlation with the correct PN sequence despreads the spread spectnlm signal and
restores the modulated message in the same nanow band as the odginal data, rvhereas crossconelating the signal from an undesired user results in a very small amount of lide band
noise at the receiver output.

Spread Spectrum: Spread spectrum is a means of transmission in rvhich the signal occupies
of the minimum necessary to send the information; the band spread is
accomplished by means of a code that is independent of the data, and a synchronized
reception r.vith the code at the receiver is used for dispreading and subsequent data recovery.

a band*'idth in excess

There are two t)?es of spread spectrum.

o
r

Direct-Sequence Spread Spectrum (DS-SS), which is employed in the lS-95 standard.

This principle is discussed in this article.
Frequency-llopping Spread Spectmm (FH-SS), where the carrier frequency of the
signal is moved (hopped) around in the band in a pseudorandom fashion.'fhe result is
an increase in effective bandr.vidth over time.

Dircct Sequence Spread Spectmm @S-SS): A direct sequence spread spectmm (DS-SS)
systcrr spreads the baseband data by directly multiplying the baseband data pulses rvith a
too

Pfcnrrc'd b)

- l,rol.

pseuoo-notse scquence that is prodtrced

by a pseudo_noise code
symbot of rhe pN rvaveform is'called.a
chip. Ths
'
dlrgrarn ol-a DS systenr rvith binary phase
rirodulatioi. " "

i;il;fi;::i"h_*

ir.ld.

Sailirr Rahnrn

"tl:i:::ii,::Ji

Transm ined
S

ignal

Figure : Block diagram ofa DS_SS system

transmitter rvith binary phase nodulation

Tle received

spread spectrum signal for a singlc user

can be represented as

l2E

"*(/) =,/ a n,(t p(t) coslzEf,t +0)

is the data sequence, p/) is the pN spreading
sequence,.f,
I

where m(t1

is the carrier phase angle ar

l=0. The autu

rectangular prtr.s, .u.r.,'oi*r,i.t r,a. an

*uu.for*i, uii.,

is the canier frequency,

;i :::51'^:iT,;ffi,lii:i,i

represents a data symbor and has duration "rpri,"J" "qr.i,;

7,. Each puise in p(ry represents a chip. is
usuatv
rectangular with an amplirude equal to +l
or _1, und hu, u arriti""
Zi.
the data symbols and chips coincide such
ttrat ttr" ratio
to.-lr" i, "i
on integer, lf Bss

il,;;;;";j,;;;;;i

i,

is the
bandwidth of s"(t) and B is the bandrvidth
oi--o
modulated signal
n(t)cos(znfcr), the spreading due to p(t) gives B* >>
"tn"u"n,,onalry
-8.

Figure below shorvs the received. spectra

of the desired spectrum
'uysrgnar and the interrercncc
at the output of the receiver wideband.firter.
uurtrf rf,atio"
the spreading rvavefomr
produces the spectra of Figure (b) at the
demodurator ;;";. ih. srgnar bandwidth
is reduced
to B' whiie the interference energy is rpr.ud ou.,
ni-u""a.,"io,rr exceeding B... The
filtering action of the demodurator removes most of "r''
the interference spectrum that does not

overlap rvith the signal spectrum

Spectral

Spectral

Density

Dens

Frequency

ity

Frequency

(a)

(b)
200

I
Prcparcd br

- prof Md

Sarfrrr Rahmar

Figure : Spectra of
desircd recer
wilh interlercnce: (a)
wideband firrer ouput
lis"l
rul .,"I"to,,
and
outpui."ft;;il;::"J,i?
"Jll9

h,c _^-, ^r .,

il:',i:i lgfi:'fi.:::.^y ;. ";;;;':, spreading ona n,in,n,u,u

;,
"#"T:ji::;:i|i
:l; ii
""" r,i,h "," ""6::il, il, :f Tffi i [ il:T:,:",""lfi
l*.r,f"1".J;."]:i
ca

pa

b i J i ry

i,

PG

*,r1::F:.*rrocess

=:=x. = 8,,

ing gain or the,r,,"",',

l;"

o11,".

'',,,

be its ability to suppress

in-

fr:l:"x? J:3fl.*,:x.aH

xn*".uffi *'*t*#*u'::l,;*m$+;.f
*"..1

,""rrI

Figure
F.ach

Stot

Stor 2

Slor 3

,An example offrequency

hopping system beflveen
vlLrrssrr J
3 cnannels.
channels.

chrnnel is defined

AS

,,'; ;

.x

il;;
iilglto[]:,cJ
F, f

a sne.rrar ro-:^_ ..,

,he hopse, and

il ilil":'}J:,*:},;",,::,,::l,j,lfrequency,n
.j
TiHfi
I"T
ff
i ffi l Tl.l
*:[li
ff $.,.' ffi1
i:.86:I,'l'#f,":",'*
ffi :lT::ff::lli## jdil:[,'";m"m;",'.*i:::::
J"f :'"1*:.*r#,,t""#{'":',#:H,;li,,,i[l[ii',,'"Jt]TJ;,n*
i:j[]"ffi
',;{{i
':Jilil,'f '.;ff
iil'['.lil::ii:ff
"H1f#jt+tr*ii,11;LT::t::,y','"Hilfti
i::i':fl.lJ,i;;;i:,#illj
gl *.r, "r,unn.i,
i;iili.:i'".1#i';*11'-?:il';T,lHi;:*:i.l:j
F::tiH'J.Tfi
flll]li:
are sent
modutarion
before rhe

fr

-un#tr.,r,"^

[i

il;l

using conventional
nano[,band

$,.$#*;ffi **ffi#**g-.;3ffi *
201

Prcpared by

Prof. i!ld. Saifur Rahman

Itlodu lator

Figure : Block diagram of frequency hopping (FH) system with single channer
modulalion.

CDMA using Direct Sequence Spread Spectrum

(DS_SS) Technique

Traditional rvays of separating signals in time (i.e. tirne division murtiple acccss,
TDMA). or
in frequency (i.e. frequency division multiple access, FDMA) are rcratively simpre u.a1s
of
making sure that the signals are orthogonal and noninterfering.
Hou'ever, in cDMA different users occupy the same bandrvidth at
the same time, but are
scparated from each other via the use of a set of orthogonar waveforms,
sequences, ;,
Here' at first, rve summarize the properties of the set oi orthogonar codes
to !e used in DS-SS

.;d.,

multiple access:

l. The cross-correlation should

be zero or very small.

Each sequence in the set has an equar number of ls and -ls, or the number of ls
differs lrorn the number of-ls by at most L
3. The scalcd dot product ofeach code should be equal to l.

Note;

* The second property gives

that particular code the pseudorandom nature.
The third property requires that the dot product of each code scaled by the order
of the code must
equal to l. The order ofthe code is effectively the length ofthe code.

**

Trvo real-valued rvaveforms x and y are said to be orthogonal if their cross-correration

R.r(0)
o\ er T is zero' rvhere
R",(0r .- alt4t) y(t1 dt

In discrete time, the two sequences

zeto,
The cross product is defined as

and

R,r(0)
where

are orthogonal

: xry =lLtt

if their cross-products R*\(0) is

xiyl

xr = [x1 x2 ... x1]

Y' = [Yr Y2 ... yL]
Notc that xr or yl clenotes the transposc ofthe column vector, which is another representation
of a sequence of nLrnrbers. For example, the following two sequences or codes, x and y, are

onhogonal:

x,

=1_t _t

y1

=;-t tt-t1

11

202

-,1

Prcpared

b)'-

Prof. Nld Srrfur

because their cross-conelation is zero: that is.

R*r(O) = xr y = (-l X-l

)+(-lXl

)+( l

Xl)+(lX- l)=0

The dot product is defined as a scalar obtained by multiplying the sequence by itself and
summing the individual terms; that is, the dot product of the code x is

R*(q=xrx=ti=t

xi xi

The two orthogonal codes in the previous example also satisi/ the second and the third
conditions. Both x and y have an equal number of ls and -ls and the scaled dot products are
xr x)t 4 =

(- I X- I )+(- I X- I )+( I )( I )+( I X I )=4/4=

(yr y)t 4=

(- I )(- I )+( I X I )+( 1 X i )+(- I X- I )=4/4=

Note that the order of each code is 4.

Operating principle of a DS-SS CDMA Multiple Access Scheme: The follou'ing Figure
illustrates the principle of DS-SS multiple access scheme. Although these systems are often
used for digital communication, we have shown their continuous-time equivalent in order to
illustrate operating principles. We show two users simultaneouslv transmitting two separate
mcssages, m1(t) and m2(t), in the same frequency band at the same time

m,(t)c,(r)

r)

m,(r) cr(r)

Transminers

Channel

Receive r

Figure : An exanple showing thc operating principle of DS-SS multiplc access. Two users are sending two
scparate messages, m r(t) and mr(t), simultaneously through the same channel in the same frequency band and at
the same time. Through the use oforthogonal codes c1(t) and c1(t), the receiver rccovers the two messages
perfectly.

o
o
.

The two users are separated from each other via the multiplication of orthogonal
codes c1(t) and c2(t), which are the continuous-time versions of the t$'o orthogonal
codes x and y mentioned previously.
Messages m1(t) is multiplied by the code c1(t), and messaBe mz(t) is multiplied by the
code c2(t).

The resulting products are added together by the adder and transmitted through the
channel. In this case, we assume perfect synchronization of the codes at the receiver.
If there are negligible enors over the channel, the recovered messages ml(t) and mr(t)
will match the original messages m1(t) and mdt) perfectly.
203

Preparcd by

ln this example, we are interested in sending two separate messages:

and m2, which is (+1,+1,-1).

- Prof Nld. Sarfur Rahman

which is (+ I ,- I ,+ I ),

the
The next Figure shows the waveforms and spectra fol the two messages ml(t) and nlz(0,
trvo orthogo-nal codes cl(t) and c2(t), and the two spread messages m1(t)c1(t) and mdt)clt)'

It should be noted here that the bandwidth of a random digital waveform is limited to l/T,
where T is the bit interval of the random digital waveform. we further make the distinction

chip
betrveen Tr and T", where Tb is the bit interval (in seconds) of the message and T" is the
interval ofthe running orthogonal code.
rate
ln this example, the chip rate (1/T") of the orthogonal code is running at four tines the bit
(l/T6). Therifore, we have an effictive bandwidth expansion factor of four. Thc bandrvidth
i*p#rion factor is sometimes called the processing gain-or (W/R), where W is the final
Note that
bandwidth of the spread message and R is the bandwidth of the baseband message
processing gain
in this example, w is equivaleit to (l/T"), R is equivalent to (l/Tr), and the
(WR) is 4, or 6 dB. (Note that Tu=4Tc i.e., T6/T"=4)

Note that after spreading by the orthogonal codes, the spread messages mr(t)cr(t) and
m:(t)c:(t) now occupy a larger bandwidth than the original messages

l)rcprrcd

by l,rol ilftj Sartirr Rrhmai

n,\!)
M,(t)

c,(t)

i l
tdt-

T.

m,(r)c,(1)

M\(tlc,(t)

m.(l)

M,(fYC,U)

1tT,

1/7,

Figure : Time waveforms

and frcquency specka, for the basband

messages mr(r) and m:(t). orlhogonal
codes
cr(r) and cr(t), and spread messages mr(r)cr(t)

an;mr0)c:(t).

Thc next Figure shorvs the waveforms at

different points
A is thc result of the summation of the trvo ,pr"^J;;;;;. of the recerver. The signal at point
,n-"
ar A now conrains
rwo separate signars rn order to recover
"0..r*.tiom
iha t*o ,"pu'.ut. messages
the composite
the signal ar A is multiplied by rhe rwo
..rp."tiu. orthogl""i ;.;';";;;;";;

:::.#.

205

Prcpared

by- Prof lvd

Sarlirr Rahmrn

Figure : Time waveforrns and frequency spectra for the signals at different points of the
receiver.

The follorving Figure shows the signals at Cr and C2, the outputs ml and m2 of the decision
thresholds, and the recovered messages ml(t) and m2(t).

The integrator adds up the signal power over one bit interval T6 of the baseband message' and
the deciiion threshold decides, based on the output of the integator, whether or not the
particular bit is a +1 or-l.Ifthe output ofthe integrator is. greate! than 0, then the decision is
a +t; if the integrator output is less than 0, then the decision is a -1. The digital-to-analog
(D/A) converter iransforms the decision into the recovered waveforms ml(t) and mr(t)'

As one can see in this idealized example, the recovered messages m1(t) and m2(t) match
perfectly the original baseband messages ml(t) and m2(t).
Problems of CDMA
DS-SS
The example presented in the previous section only serves to illustlate the principle of

multiple access. We have just demonstrated that, using DS-Ss techniques' separate messages
and lhe
can be sent through the same channel in the same ftequency band at the same time'
world
real
are many
messages can be iuccessfully recoverecl at the receiver' However, there
phenotena, especially in a mobile communications environment, that degradc the
performance of such a DS-ss multiple access system. Thre ale two problcms: thc ear-far
problem, and the partial correlation problem
206

Prcparcd by

ml

Prof. Md Saifur Rahnran

+l

n,(t)

lrt,o

Figure : Tirne wavefofins at the output of the integrators and decision threshold.
ln mobile communications, each user is geographically dispersed but transmitting in the same
frequency spectrum using DS-sS. Some users are closer to thc base station thln others.
Thc
result is that powers received from those users that are close by are higher than powers
received from users that are farther away. Because all users are fansmitting in the saml
band,
the higher received powers from users nearby constitute an interference that desrades the
general performance of the system. In order to combat this near-far phenomenin, power
control is utilized to make sure that the powers received at the base statlon are the same for
all users ln the earlier example, we have assumed perfect power control by specifying that
mr(t)cr(t) and m2(t)cdt) both have the same amplitudes (i.e. ranging fiom +l to _l). ' The second problem is pdrt.al correlatiort. The problem comes rvhen there is no anemDt to
synchronize the transmitters sharing the same band. Even rvhen the rransmitters are
synchronized, there is still the problem of propagation delay, rvhich is inherent in a nobile
channel. For example, the trvo codes mentioned previously are orthogonal rvhen they arc
perfectly aligned:

-l
-1

-1
+l

+1 +l
+t -l

-1
-1

-l
+l

,fl
+l

+1
_1

201

I'rt-p.rrcd

Ilorvcvcr,

if yi suffcrs a dclay of

br

Prof irl,l

rrfur P.rhtnrn

one chip (ir:fl shifte b;'one chip dtrntion) as it rcsult

of

propagation delay in a mobiie channel, tllen

f ,.t

-t
+l

-l +l
+1 -1

+l
-1

-1
+i

-l
+l

+1
-1

+i

-l

verify that these two sequences ale no longer orthogonal. If the codes are
not orthogonal due to synchronization or channel impairment, then multiple access messages
in the same band can no longer be separated fiom one another via code orthogonality. The
fssults are conelation crosstalk and mutual interference. ln essence, an additional condition
Readers can easily

necds to be adhered to: that

is'

' r(rr r(l r rlr/r

=0

x(t)IQ+r-T)dt-0
Therefore, simple orthogonality between trvo aligrred codes is not enough the above trvo
partial correlations must also be zero, or at least small, for any valLIe of r Iikelr encotlntered
in the s),stem.

Applications of DS-SS in Mobilc c.ommunication: Despite its difficulties, rvhich are easill'
solved rvith optimized systen design, CDMA does have its adlantages lvhen applied to
mobile communications.

First of all, a GDMA system can readily take advantage of the 'lorce oclil,i4' of nonlll
human spccch. In a twr-r-pcrson convcrslticrn, clch spcakcr is activq lcss than lrllf of thc tintc
During the quiet pcriod, the transmitiels could effectively turn off and reducc interfcrence
po*"iint odu""d into the channel. This reduction in interference can translale into crpllcil)
gain for the system.
The second advantage is that in CDMA, the physical RF channel can be reused in everl'cell.
thus giving a freqrtency reuse factor ofclose to 1. Therc is no necd to frequenncy plan in
CDMA, which may be one of the welcoming benefits for RF design engineers

The third advantage is CDlvlA's ability to mitigate multipath distortion'

I;urthcrmore, a cDMA system can take advantage of multipaths by using thc
rvhich demodulates and uses the sigrral energy of all paths.

208

rlkc

rt'ccir crs'