The Data Link Layer

Data Link layer

Link Layer services
• Encoding
• encode binary data into electromagnetic signals
• Framing
• encapsulate data into frame, adding header, trailer
‘physical addresses’ used in frame headers to identify
• source, dest (different from IP address!)
• Link access control
• coordinate access for shared link
• Flow control
• pacing between sending and receiving nodes such that the
sender does not overwhelm the receiver
• nodes on each side of a link have a limited amount of packet buffering
capacity
 Types of errors
• An error occurs when a bit • Single bit error
is altered between
transmission and
reception
• Binary 1 is transmitted and • Bit error – 2 or more bits changed
binary 0 is received
• Binary 0 is transmitted and
binary 1 is received
Error detection process
 Parity checking
• Single bit parity: • Two Dimension Bit Parity
Detects up to 3 bit errors
Detects single bit errors

Read about
• Internet checksum
• Cyclic Redundancy Check
Link Layer Multiple Access Protocols
two types of “links”:
• point-to-point
• PPP for dial-up access
• point-to-point link between Ethernet switch, host
• broadcast (shared wire or medium)
• Older Ethernet
• upstream Hybrid fiber-coaxial (HFC)
• 802.11 wireless LAN
 Multiple Access Protocol
• Multiple access protocol:
• distributed algorithm that determines how nodes share channel, i.e.,
• determine when node can transmit
• communication about channel sharing must use channel itself!
• Ideal multiple access protocol:
For a multiple access channel of rate R bps
1. When one node wants to transmit, it can send at rate R.
2. When M nodes want to transmit, each can send at average rate
R/M.
3. Fully decentralized:
• no special node to coordinate transmissions
• no synchronized clocks
• fault-tolerant/robust
Types of Multiple access Protocols
• Three broad classes:
• Channel partitioning
• divide channel into smaller “pieces” (time slots, frequency, code)
• allocate piece to node for exclusive use
• Random access
• channel not divided, allow collisions
• “recover” from collisions
• Centrally controlled/coordinated
• tightly coordinate shared access to avoid collisions
Channel partitioning
• TDMA: Time Division Multiple Access
• access to channel in "rounds"
• each station gets fixed length slot (length = pkt trans time) in
each round
• unused slots go idle

• FDMA: frequency division multiple access

• channel spectrum divided into frequency bands
• each station assigned fixed frequency band
• unused transmission time in frequency bands go idle
Channel portioning cont’d
CDMA: Code division multiple access
• Transmissions are combined on the same channel at the same
time but are separated by codes
• Uses coding to allow multiple simultaneous transmission
without interference.
• Each user has it’s unique code.
 Channel portioning cont’d

• F/TDMA
• 2G cellular
network

CDMA
3G cellular
network
How good are channel portioning
protocols
• Broadcast channel of rate R bps
1. When all M nodes want to transmit each can send at average
rate R/M
2. When one node wants to transmit it can send at rate R/M.
Inefficient!
3. Decentralization/fault tolerance:
• no special node to coordinate transmissions
• synchronized clocks for all but FDMA
Random Access protocols
• When node has data to send
• transmit at full channel data rate R.
• no a priori coordination among nodes
• Two or more transmitting nodes -> “collision”,
• Random access protocol specifies:
• how to detect collisions
• how to recover from collisions (e.g., via delayed retransmissions)
• Examples of random access protocols:
• slotted ALOHA
• ALOHA
• CSMA, CSMA/CD
CSMA/CD
Used in ethernet

After
If a collision transmittin
If the is detected, g the
If the medium is transmit a jamming
medium busy, brief signal,
jamming wait a
is idle, continue to signal to random
transmit; listen until assure that amount of
otherwis the channel all stations time,
e, go to is idle, then know that referred to
step 2 transmit there has as the
been a backoff,
immediately collision and then
cease attempt to
transmission transmit
again
Ethernet
• dominant” for local-area networks:
• Developed at Xerox PARC in 1970s
• First widely used LAN technology
• Random access multiple access control

• Ethernet Frame structure

• Sending adapter encapsulates data payload in Ethernet frame
Ethernet frame
• Preamble:
• 7 bytes with pattern 10101010 followed by one byte with pattern10101011
• indicating the beginning of a frame, synchronizing receiver/sender clocks
• Addresses:
• link layer addresses, different from IP addresses
• Length: indicates the length of the body field (max 1500bytes)
• CRC: checked at receiver
• if error is detected, the frame is simply dropped
Connectionless, Unreliable
• Connectionless:
• No handshaking between sending and receiving adapter.
• Unreliable:
• corrupted frames are simply dropped
• no attempt is made on the receiver to detect and recover lost frames
• corrupted or lost packets may be recovered by higher layer protocols,
e.g. TCP
Ethernet uses CDMA/CD
• Random access, no • Before attempting a
synchronized clocks retransmission, adapter waits a
• adapter doesn’t transmit if it random
senses that some other
adapter is transmitting, that is,
carrier sense
• transmitting adapter aborts
when it senses that another
adapter is transmitting, that is,
collision detection
How good are random access protocols
• Multiple channel of rate R bps
1. When one node wants to transmit,
• it can send at full rate R.
2. When M nodes want to transmit,
• each can send at average rate less than R/M.
Not ideal!
3. Decentralization:
• no special node to coordinate transmissions
• no synchronized clocks except for slotted ALOHA
Centrally Controlled MAC protocols
• Polling: • Token passing:
• master node “invites”
• control token passed from one
• slave nodes to transmit in turn
• node to next sequentially.
• concerns:
• polling overhead (bandwidth & latency)
• transmit only when holding the token
• single point of failure (master) • concerns:
• token overhead (bandwidth & latency)
• single point of failure (token)
LAN Interconnection
• MAC addresses and ARP
• MAC (or LAN or physical or Ethernet) address:
• function: used ‘locally” to get frame from one interface to another
physically-connected interface (same network, in IP-addressing sense)
• 48 bit MAC address (for most LANs) burned in NIC ROM, also
sometimes software settable
• e.g.: 1A-2F-BB-76-09-AD
• Universally unique
Question
• How do we determine node
A’s link layer address knowing
it’s IP address.?
Address Resolution Protocol (ARP)
• Specific to internet Architecture
• Each node (Host, Router) on a LAN has an ARP table
• ARP Table: IP/Link layer address mappings for some LAN
nodes
 How ARP work?
• A wants to send datagram to B with an IP
address.
• Suppose B’s link-layer address is not in
A’s ARP table.
• A broadcasts ARP query packet,
containing B's IP address.
• B receives ARP packet, replies to A with
its (B's) link-layer address.
• A caches (saves) IP-to-link layer address
pair in its ARP table until it times out when
can a mapping change?
• Broadcasting an ARP request • ARP reply
Interconnecting Nodes in LAN
• Hubs: physical-layer signal repeaters.
• Bridges: understands link-layer protocol (Ethernet), smarter
than hubs.
• Switches: essentially bridges with large number of ports.

• What is the difference between bridges, switches and hubs?

Questions ?