Chapter 5: The Data Link Layer

Our goals:
 understand principles behind data link layer
services:
 error detection, correction
 sharing a broadcast channel: multiple access
 link layer addressing
 reliable data transfer, flow control: done!
 instantiation and implementation of various link
layer technologies

5: DataLink Layer 5-1

Link Layer: Introduction “link”
Some terminology:
 hosts and routers are nodes
 communication channels that
connect adjacent nodes along
communication path are links
 wired links
 wireless links
 LANs
 layer-2 packet is a frame,
encapsulates datagram

data-link layer has responsibility of

transferring datagram from one node
to adjacent node over a link
5: DataLink Layer 5-2
Link layer: context
 Datagram transferred by
transportation analogy
 trip from Princeton to
different link protocols
Lausanne
over different links:
 limo: Princeton to JFK
 e.g., Ethernet on first link,
 plane: JFK to Geneva
frame relay on
intermediate links, 802.11  train: Geneva to Lausanne
on last link  tourist = datagram
 Each link protocol  transport segment =
provides different communication link
services  transportation mode =
e.g., may or may not

link layer protocol
provide rdt over link
 travel agent = routing
algorithm
5: DataLink Layer 5-3
Link Layer Services
 Framing, link access:
 encapsulate datagram into frame, adding header, trailer
 channel access if shared medium
 “MAC” addresses used in frame headers to identify
source, dest
• different from IP address!
 Reliable delivery between adjacent nodes
 we learned how to do this already (chapter 3)!
 seldom used on low bit error link (fiber, some twisted
pair)
 wireless links: high error rates
• Q: why both link-level and end-end reliability?

5: DataLink Layer 5-4

Link Layer Services (more)
 Flow Control:
 pacing between adjacent sending and receiving nodes
 Error Detection:
 errors caused by signal attenuation, noise.
 receiver detects presence of errors:
• signals sender for retransmission or drops frame
 Error Correction:
 receiver identifies and corrects bit error(s) without
resorting to retransmission
 Half-duplex and full-duplex
 with half duplex, nodes at both ends of link can transmit,
but not at same time
5: DataLink Layer 5-5
 Adapters Communicating
datagram
link layer protocol rcving
sending node
node
frame frame
adapter adapter

 link layer implemented in  receiving side

“adapter” (aka NIC)  looks for errors, rdt, flow
 Ethernet card, PCMCIA control, etc
card, 802.11 card  extracts datagram, passes
to rcving node
 sending side:
 encapsulates datagram in  adapter is semi-
a frame autonomous
 adds error checking bits,  implements link &
rdt, flow control, etc. physical layers
5: DataLink Layer 5-6
Error Detection
EDC= Error Detection and Correction bits (redundancy)
D = Data protected by error checking, may include header fields

• Error detection not 100% reliable!

• protocol may miss some errors, but rarely
• larger EDC field yields better detection and correction

5: DataLink Layer 5-7

Parity Checking
Single Bit Parity: Two Dimensional Bit Parity:
Detect single bit errors Detect and correct single bit errors

0 0

5: DataLink Layer 5-8

Internet checksum
Goal: detect “errors” (e.g., flipped bits) in transmitted
segment (note: used at transport layer only)

Sender: Receiver:
 compute checksum of received
 treat segment contents as
segment
sequence of 16-bit
 check if computed checksum
integers equals checksum field value:
 checksum: addition (1’s  NO - error detected
complement sum) of  YES - no error detected. But
segment contents maybe errors nonetheless?
 sender puts checksum More later ….
value into the checksum
field

5: DataLink Layer 5-9

Checksumming: Cyclic Redundancy Check
 view data bits, D, as a binary number
 choose r+1 bit pattern (generator), G
 goal: choose r CRC bits, R, such that
 <D,R> exactly divisible by G (modulo 2)
 receiver knows G, divides <D,R> by G. If non-zero remainder:
error detected!
 can detect all burst errors less than r+1 bits
 widely used in practice (Ethernet, 802.11, ATM, HDLC)

5: DataLink Layer 5-10

CRC Example
Want:
D.2r XOR R = nG
equivalently:
D.2r = nG XOR R
equivalently:
if we divide D.2r by
G, want remainder R

D.2r
R = remainder[ ]
G

5: DataLink Layer 5-11

Multiple Access Links and Protocols
Two types of “links”:
 point-to-point
 PPP for dial-up access
 point-to-point link between Ethernet switch and host

 broadcast (shared wire or medium)

 traditional Ethernet
 upstream HFC (hybrid fiber-coax used in cable TV)
 802.11 wireless LAN

802.11)
5: DataLink Layer 5-12
Multiple Access protocols
 single shared broadcast channel
 two or more simultaneous transmissions by nodes:
interference
 collision if node receives two or more signals at the same time
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!
 no out-of-band channel for coordination

5: DataLink Layer 5-13

Ideal Multiple Access Protocol
Broadcast 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 synchronization of clocks, slots
4. Simple

5: DataLink Layer 5-14

MAC Protocols: a taxonomy
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

 “Taking turns”
 Nodes take turns, but nodes with more to send can take
longer turns

5: DataLink Layer 5-15

Channel Partitioning MAC protocols: TDMA

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
 example: 6-station LAN, 1,3,4 have pkts, slots
2,5,6 idle

5: DataLink Layer 5-16

Channel Partitioning MAC protocols: FDMA
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
 example: 6-station LAN, 1,3,4 have pkts, frequency
bands 2,5,6 idle
frequency bands

5: DataLink Layer 5-17

Random Access Protocols
 When node has packet to send
 transmit at full channel data rate R.
 no a priori coordination among nodes

 two or more transmitting nodes → “collision”,

 random access MAC protocol specifies:
 how to detect collisions
 how to recover from collisions (e.g., via delayed
retransmissions)
 Examples of random access MAC protocols:
 slotted ALOHA
 ALOHA
 CSMA, CSMA/CD, CSMA/CA

5: DataLink Layer 5-18

 Slotted ALOHA
Assumptions Operation
 all frames same size  when node obtains fresh
 time is divided into frame, it transmits in next
equal size slots, time to slot
transmit 1 frame  if no collision, node can
 nodes start to transmit send new frame in next
frames only at beginning slot
of slots  if collision, node
 nodes are synchronized retransmits frame in each
subsequent slot with prob.
 if 2 or more nodes
transmit in slot, all p until success
nodes detect collision
5: DataLink Layer 5-19
Slotted ALOHA

Pros Cons
 single active node can  collisions, wasting slots
continuously transmit  idle slots
at full rate of channel  nodes may be able to
 highly decentralized: detect collision in less
only slots in nodes than time to transmit
packet
need to be in sync
 clock synchronization
 simple
5: DataLink Layer 5-20
Slotted Aloha efficiency
Efficiency is the long-run  For max efficiency
fraction of successful slots with N nodes, find p*
when there are many nodes, that maximizes
each with many frames to send Np(1-p)N-1
 For many nodes, take
 Suppose N nodes with limit of Np*(1-p*)N-1
many frames to send, as N goes to infinity,
each transmits in slot gives 1/e = .37
with probability p
 prob that node 1 has At best: channel
success in a slot used for useful
= p(1-p)N-1 transmissions 37%
 prob that any node has of time!
a success = Np(1-p)N-1
5: DataLink Layer 5-21
Pure (unslotted) ALOHA
 unslotted Aloha: simpler, no synchronization
 when frame first arrives
 transmit immediately

 collision probability increases:

 frame sent at t0 collides with other frames sent in [t0-1,t0+1]

5: DataLink Layer 5-22

Pure Aloha efficiency
P(success by given node) = P(node transmits) .
P(no other node transmits in [t0-1,t0] .
P(no other node transmits in [t0,t0+1]
= p . (1-p)N-1 . (1-p)N-1
= p . (1-p)2(N-1)

… choosing optimum p and then letting n → 

= 1/(2e) = .18
Even worse !

5: DataLink Layer 5-23

CSMA (Carrier Sense Multiple Access)

CSMA: listen before transmit:

If channel sensed idle: transmit entire frame
 If channel sensed busy, defer transmission

 Human analogy: don’t interrupt others!

5: DataLink Layer 5-24

 CSMA collisions spatial layout of nodes

collisions can still occur:

propagation delay means
two nodes may not hear
each other’s transmission

collision:
entire packet transmission
time wasted
note:
role of distance & propagation
delay in determining collision
probability

5: DataLink Layer 5-25

CSMA/CD (Collision Detection)
CSMA/CD: carrier sensing, deferral as in CSMA
 collisions detected within short time
 colliding transmissions aborted, reducing channel
wastage
 collision detection:
 easy in wired LANs: measure signal strengths,
compare transmitted, received signals
 difficult in wireless LANs: receiver shut off while
transmitting
 human analogy: the polite conversationalist

5: DataLink Layer 5-26

CSMA/CD collision detection

5: DataLink Layer 5-27

“Taking Turns” MAC protocols
channel partitioning MAC protocols:
 share channel efficiently and fairly at high load
 inefficient at low load: delay in channel access,
1/N bandwidth allocated even if only 1 active
node!
Random access MAC protocols
 efficient at low load: single node can fully
utilize channel
 high load: collision overhead
“taking turns” protocols
look for best of both worlds!
5: DataLink Layer 5-28
“Taking Turns” MAC protocols
Polling: Token passing:
 master node  control token passed from
“invites” slave nodes one node to next
to transmit in turn sequentially.
 concerns:  token message
 polling overhead  concerns:
 latency  token overhead
 single point of  latency
failure (master)  single point of failure (token)

5: DataLink Layer 5-29

Summary of MAC protocols
 What do you do with a shared media?
 Channel Partitioning, by time, frequency or code
• Time Division, Frequency Division
 Random partitioning (dynamic),
• ALOHA, S-ALOHA, CSMA, CSMA/CD
• carrier sensing: easy in some technologies (wire), hard
in others (wireless)
• CSMA/CD used in Ethernet
• CSMA/CA used in 802.11
 Taking Turns
• polling from a central site, token passing

5: DataLink Layer 5-30

MAC Addresses and ARP

 32-bit IP address:
 network-layer address
 used to get datagram to destination IP subnet

 MAC (or LAN or “physical” or Ethernet)

address:
 used to get datagram from one interface to
another physically-connected interface (same
network)
 48 bit MAC address (for most LANs)
burned in the adapter ROM

5: DataLink Layer 5-31

 LAN Addresses and ARP
Each adapter on LAN has unique LAN address

1A-2F-BB-76-09-AD Broadcast address =

FF-FF-FF-FF-FF-FF

LAN
(wired or = adapter
wireless)
71-65-F7-2B-08-53
58-23-D7-FA-20-B0

0C-C4-11-6F-E3-98

5: DataLink Layer 5-32

LAN Address (more)
 MAC address allocation administered by IEEE
 manufacturer buys portion of MAC address space
(to assure uniqueness)
 Analogy:
(a) MAC address: like Social Security Number
(b) IP address: like postal address
 MAC flat address allows easier portability
 can move LAN card from one LAN to another
 IP hierarchical address NOT portable
 depends on IP subnet to which node is attached

5: DataLink Layer 5-33

 ARP: Address Resolution Protocol

Question: how to determine  Each IP node (Host,

MAC address of B Router) on LAN has
knowing B’s IP address? ARP table
 ARP Table: IP/MAC
237.196.7.78
address mappings for
1A-2F-BB-76-09-AD
some LAN nodes
237.196.7.23 < IP address; MAC address; TTL>
237.196.7.14
 TTL (Time To Live): time
LAN after which address
71-65-F7-2B-08-53
58-23-D7-FA-20-B0
mapping will be forgotten
(typically 20 min)

0C-C4-11-6F-E3-98
237.196.7.88

5: DataLink Layer 5-34

ARP protocol: Same LAN (network)
 A wants to send datagram
to B, and B’s MAC address  A caches (saves) IP-to-
not in A’s ARP table. MAC address pair in its
 A broadcasts ARP query ARP table until information
packet, containing B's IP becomes old (times out)
address  soft state: information
 Dest MAC address = that times out (goes
FF-FF-FF-FF-FF-FF away) unless refreshed
 all machines on LAN  ARP is “plug-and-play”:
receive ARP query  nodes create their ARP
 B receives ARP packet, tables without
replies to A with its (B's) intervention from net
MAC address administrator
 frame sent to A’s MAC
address (unicast)

5: DataLink Layer 5-35

 Addressing: routing to another LAN
walkthrough: send datagram from A to B via R
▪ focus on addressing – at IP (datagram) and MAC layer (frame)
▪ assume A knows B’s IP address
▪ assume A knows IP address of first hop router, R (how?)
▪ assume A knows R’s MAC address (how?)

A B
R
111.111.111.111
222.222.222.222
74-29-9C-E8-FF-55
49-BD-D2-C7-56-2A
222.222.222.220
1A-23-F9-CD-06-9B

111.111.111.112 111.111.111.110 222.222.222.221

CC-49-DE-D0-AB-7D E6-E9-00-17-BB-4B 88-B2-2F-54-1A-0F

5-36
Addressing: routing to another LAN
▪ A creates IP datagram with IP source A, destination B
▪ A creates link-layer frame with R's MAC address as destination address,
frame contains A-to-B IP datagram
MAC src: 74-29-9C-E8-FF-55
MAC dest: E6-E9-00-17-BB-4B
IP src: 111.111.111.111
IP dest: 222.222.222.222

IP
Eth
Phy

A B
R
111.111.111.111
222.222.222.222
74-29-9C-E8-FF-55
49-BD-D2-C7-56-2A
222.222.222.220
1A-23-F9-CD-06-9B

111.111.111.112 111.111.111.110 222.222.222.221

CC-49-DE-D0-AB-7D E6-E9-00-17-BB-4B 88-B2-2F-54-1A-0F

5-37
Addressing: routing to another LAN
▪ frame sent from A to R
▪ frame received at R, datagram removed, passed up to IP

MAC src: 74-29-9C-E8-FF-55

MAC dest: E6-E9-00-17-BB-4B
IP src: 111.111.111.111
IP dest: 222.222.222.222
IP src: 111.111.111.111
IP dest: 222.222.222.222

IP IP
Eth Eth
Phy Phy

A B
R
111.111.111.111
222.222.222.222
74-29-9C-E8-FF-55
49-BD-D2-C7-56-2A
222.222.222.220
1A-23-F9-CD-06-9B

111.111.111.112 111.111.111.110 222.222.222.221

CC-49-DE-D0-AB-7D E6-E9-00-17-BB-4B 88-B2-2F-54-1A-0F

5-38
Addressing: routing to another LAN
▪ R forwards datagram with IP source A, destination B
▪ R creates link-layer frame with B's MAC address as destination address,
frame contains A-to-B IP datagram

MAC src: 1A-23-F9-CD-06-9B

MAC dest: 49-BD-D2-C7-56-2A
IP src: 111.111.111.111
IP dest: 222.222.222.222
IP
IP Eth
Eth Phy
Phy

A B
R
111.111.111.111
222.222.222.222
74-29-9C-E8-FF-55
49-BD-D2-C7-56-2A
222.222.222.220
1A-23-F9-CD-06-9B

111.111.111.112 111.111.111.110 222.222.222.221

CC-49-DE-D0-AB-7D E6-E9-00-17-BB-4B 88-B2-2F-54-1A-0F

5-39
Addressing: routing to another LAN
▪ R forwards datagram with IP source A, destination B
▪ R creates link-layer frame with B's MAC address as destination address,
frame contains A-to-B IP datagram

MAC src: 1A-23-F9-CD-06-9B

MAC dest: 49-BD-D2-C7-56-2A
IP src: 111.111.111.111
IP dest: 222.222.222.222
IP
IP Eth
Eth Phy
Phy

A B
R
111.111.111.111
222.222.222.222
74-29-9C-E8-FF-55
49-BD-D2-C7-56-2A
222.222.222.220
1A-23-F9-CD-06-9B

111.111.111.112 111.111.111.110 222.222.222.221

CC-49-DE-D0-AB-7D E6-E9-00-17-BB-4B 88-B2-2F-54-1A-0F

5-40
Addressing: routing to another LAN
▪ R forwards datagram with IP source A, destination B
▪ R creates link-layer frame with B's MAC address as dest, frame contains
A-to-B IP datagram
MAC src: 1A-23-F9-CD-06-9B
MAC dest: 49-BD-D2-C7-56-2A
IP src: 111.111.111.111
IP dest: 222.222.222.222

IP
Eth
Phy

A B
R
111.111.111.111
222.222.222.222
74-29-9C-E8-FF-55
49-BD-D2-C7-56-2A
222.222.222.220
1A-23-F9-CD-06-9B

111.111.111.112 111.111.111.110 222.222.222.221

CC-49-DE-D0-AB-7D E6-E9-00-17-BB-4B 88-B2-2F-54-1A-0F

5-41
Ethernet
“dominant” wired LAN technology:
 cheap: $20 for 100Mbs!
 first widely used LAN technology
 Simpler, cheaper than token LANs and ATM
 Kept up with speed race: 10 Mbps – 10 Gbps

Metcalfe’s Ethernet
sketch

5: DataLink Layer 5-42

Ethernet: physical topology
▪ bus: popular through mid 90s
• all nodes in same collision domain (can collide with each
other)
▪ star: prevails today
• active switch in center
• each “spoke” runs a (separate) Ethernet protocol (nodes
do not collide with each other)

switch
star
bus: coaxial cable
5-43
Ethernet Frame Structure
Sending adapter encapsulates IP datagram (or other
network layer protocol packet) in Ethernet frame

Preamble:
 7 bytes with pattern 10101010 followed by one
byte with pattern 10101011
 used to synchronize receiver, sender clock rates

5: DataLink Layer 5-44

Ethernet Frame Structure
(more)
 Addresses: 6 bytes
 if adapter receives frame with matching destination
address, or with broadcast address (eg ARP packet), it
passes data in frame to net-layer protocol
 otherwise, adapter discards frame

 Type: 2 bytes, indicates the higher (network)

layer protocol (commonly IP, but may also be ARP,
Novell IPX and AppleTalk, etc.)
 CRC: 4 bytes, checked at receiver, if error is
detected, the frame is simply dropped

5: DataLink Layer 5-45

Unreliable, connectionless service
 Connectionless: No handshaking between sending
and receiving adapter.
 Unreliable: receiving adapter doesn’t send acks or
nacks to sending adapter
 stream of datagrams passed to network layer can have
gaps
 gaps will be filled if app is using TCP
 otherwise, app will see the gaps

5: DataLink Layer 5-46

802.3 Ethernet standards: link &
physical layers
many different Ethernet standards
common MAC protocol and frame format
different speeds: 2 Mbps, 10 Mbps, 100 Mbps, 1Gbps, 10
Gbps, 40 Gbps
different physical layer media: fiber, cable

MAC protocol
application and frame format
transport
network 100BASE-TX 100BASE-T2 100BASE-FX
link 100BASE-T4 100BASE-SX 100BASE-BX
physical

copper (twister fiber physical layer

pair) physical layer 5-47
Ethernet uses CSMA/CD
 No slots  Before attempting a
 adapter doesn’t transmit retransmission,
if it senses that some adapter waits a
other adapter is random time, that is,
transmitting, that is, random access
carrier sense
 transmitting adapter
aborts when it senses
that another adapter is
transmitting, that is,
collision detection

5: DataLink Layer 5-48

 Ethernet CSMA/CD algorithm
1. Adapter receives 4. If adapter detects
datagram from net layer & another transmission while
creates frame transmitting, aborts and
2. If adapter senses channel sends 48-bit jam signal
idle, it starts to transmit 5. After aborting, adapter
frame. If it senses enters exponential
channel busy, waits until backoff: after the mth
channel idle and then collision, adapter chooses
transmits a K at random from
3. If adapter transmits {0,1,2,…,2m-1}. Adapter
entire frame without waits K·512 bit times and
detecting another returns to Step 2
transmission, the adapter
is done with frame ! 5: DataLink Layer 5-49
 Frame Size limitations for Ethernet
For proper collision detection: A’s  Minimum Frame size (Fmin)
frame should last at least until For proper collision detection:
B’s frame reaches A
Fmin = Min. frame size
A d B
R = Ethernet’s transmission rate,
e.g., 10 Mb/s
dmax = max. Ethernet segment
2d/S A’s frame length
S = Propagation speed (2x108
time

B’s frame m/s)

Fmin/R  2dmax/S
 Maximum Frame Size (Fmax)
For fairness among competing
A’s frame in yellow nodes
B’s frame in green Fmin=64 Bytes, Fmax=1500 Bytes

5: DataLink Layer 5-50

Ethernet’s CSMA/CD (more)
Jam Signal: make sure all Exponential Backoff:
other transmitters are  Goal: adapt retransmission
aware of collision; 48 bits attempts to estimated current
Random retransmission delay: load
K· 512 bit transmission  heavy load: random wait will be
times where K is randomly longer
selected; bit time is 0.1  first collision: choose K from
microsec for 10 Mbps and {0,1}; delay is K· 512 bit
0.01 microsec for 100 Mbps transmission times
Ethernet  after second collision: choose K
from {0,1,2,3}…
 after ten collisions, choose K
from {0,1,2,3,4,…,1023}
 for max value of K=1023: wait
See/interact with Java time is about 50 msec for 10
applet on AWL Web site: Mbps, 5 msec for 100 Mbps
highly recommended ! Ethernet
5: DataLink Layer 5-51
CSMA/CD efficiency
 tprop = max prop between 2 nodes in LAN
 ttrans = time to transmit max-size frame

1
e
f
fi
cie
nc
y
+
15t
pro
p/t
t
ran
s

 Efficiency goes to 1 as tprop goes to 0

 Goes to 1 as ttrans goes to infinity
 Much better than ALOHA, but still decentralized,
simple, and cheap

5: DataLink Layer 5-52

Ethernet switch
 link-layer device: takes an active role
 store, forward Ethernet frames
 examine incoming frame’s MAC address,
selectively forward frame to one-or-more
outgoing links when frame is to be forwarded on
segment, uses CSMA/CD to access segment
 transparent
 hosts are unaware of presence of switches
 plug-and-play, self-learning
 switches do not need to be configured

Link Layer 5-53

Switch: multiple simultaneous
transmissions
 hosts have dedicated, A
direct connection to switch
 switches buffer packets C’ B

 Ethernet protocol used on 6 1 2

each incoming link, but no
collisions; full duplex 5 4 3
 each link is its own
B’ C
collision domain
 switching: A-to-A’ and B-
to-B’ can transmit A’
simultaneously, without switch with six interfaces
collisions (1,2,3,4,5,6)

Link Layer 5-54

Switch forwarding table
Q: how does switch know A’ A
reachable via interface 4, B’
C’ B
reachable via interface 5?
❖ A: each switch has a switch table, 6 1 2
each entry:
5 4 3
▪ (MAC address of host, interface to
reach host, time stamp) B’ C

▪ looks like a routing table!

A’
Q: how are entries created, maintained switch with six interfaces
in switch table? (1,2,3,4,5,6)
▪ something like a routing protocol?

Link Layer 5-55

Switch: self-learning Source: A
Dest: A’

A A A’
 switch learns which
hosts can be reached C’ B
through which
interfaces 6 1 2

 when frame 4
5 3
received, switch
“learns” location of B’ C

sender: incoming
LAN segment A’
 records
MAC addr interface TTL
sender/location pair Switch table
A 1 60
in switch table (initially empty)

Link Layer 5-56

Switch: frame
filtering/forwarding
when frame received at switch:
1. record incoming link, MAC address of sending host
2. index switch table using MAC destination address
3. if entry found for destination
then {
if destination on segment from which frame arrived
then drop frame
else forward frame on interface indicated by entry
}
else flood /* forward on all interfaces except arriving
interface */

Link Layer 5-57

Self-learning, forwarding: Source: A
example Dest: A’

A A A’
 frame destination, A’,
locaton unknown: flood C’ B

❖ destination A location 6 1 2

known: selectively send A A’

5 4 3
on just one link B’ C
A’ A

A’

MAC addr interface TTL

A 1 60 switch table
A’ 4 60 (initially empty)

Link Layer 5-58

Interconnecting switches
 switches can be connected together
S4

S1
S3
A S2
F
D I
B C
G H
E

Q: sending from A to G - how does S1 know to

forward frame destined to F via S4 and S3?
❖A: self learning! (works exactly the same as in
single-switch case!)
Link Layer 5-59
Self-learning multi-switch
example
Suppose C sends frame to I, I responds to C

S4

S1
S3
A S2
F
D I
B C
G H
E

❖ Q: show switch tables and packet forwarding in S1, S2, S3, S4

Link Layer 5-60

Switches: dedicated access
 Switch with many A
interfaces
 Hosts have direct C’ B
connection to switch
 No collisions; full duplex
switch

Switching: A-to-A’ and B-to-B’

simultaneously, no collisions C

B’ A’
 combinations of
shared/dedicated,
10/100/1000 Mbps
interfaces possible
5: DataLink Layer 5-61
Institutional network

mail server
to external
network
router web server

IP subnet

Link Layer 5-62

Switches vs.
routers application
both are store-and-forward: transport
datagram network
▪routers: network-layer frame link
devices (examine network- physical link frame
layer headers) physical
▪switches: link-layer devices
(examine link-layer headers)
switch

both have forwarding tables: network datagram

link
▪routers: compute tables frame
using routing algorithms, IP physical
addresses application
▪switches: learn forwarding transport
table using flooding, network
learning, MAC addresses link
physical

Link Layer 5-63

Summary comparison

routers switches

traffic yes yes

isolation
plug & play no yes

optimal yes no
routing

5: DataLink Layer 5-64

 A day in the life: scenario

browser DNS server

Bilkent’s NSP
network

Bilkent network

web page

web server Google’s network

5: DataLink Layer 5-65

A day in the life… connecting to the Internet
DHCP DHCP connecting laptop needs to
UDP
get its own IP address,
DHCP
DHCP IP
DHCP Eth addr of first-hop router,
Phy addr of DNS server: use
DHCP
DHCP
 DHCP request encapsulated
DHCP DHCP in UDP, encapsulated in IP,
DHCP UDP encapsulated in 802.1
DHCP IP Ethernet
DHCP Eth
Phy
router  Ethernet frame broadcast
(runs DHCP) (dest: FFFFFFFFFFFF) on LAN,
received at router running
DHCP server

 Ethernet demux’ed to IP
demux’ed, UDP demux’ed to
DHCP

5: DataLink Layer 5-66

A day in the life… connecting to the Internet
DHCP DHCP DHCP server formulates
DHCP UDP DHCP ACK containing
DHCP IP client’s IP address, IP
DHCP Eth address of first-hop
Phy router for client, name &
IP address of DNS server
 encapsulation at DHCP
server, frame forwarded
DHCP DHCP (switch learning) through
DHCP UDP LAN, demultiplexing at
DHCP IP client
DHCP Eth
Phy
router  DHCP client receives DHCP
DHCP
(runs DHCP) ACK reply

Client now has IP address, knows name & addr of DNS

server, IP address of its first-hop router

5: DataLink Layer 5-67

A day in the life… ARP (before DNS, before HTTP)
DNS DNS before sending HTTP request, need
DNS UDP IP address of www.google.com:
DNS
ARP
IP DNS
ARP query Eth
Phy  DNS query created, encapsulated
in UDP, encapsulated in IP,
encasulated in Eth. In order to
ARP
send frame to router, need MAC
ARP reply Eth address of router interface: ARP
Phy
 ARP query broadcast, received
by router, which replies with
ARP reply giving MAC address
of router interface
 client now knows MAC address
of first hop router, so can now
send frame containing DNS
query

5: DataLink Layer 5-68

A day in the life… using DNS DNS
DNS UDP DNS server
DNS IP
DNS DNS DNS Eth
DNS UDP DNS Phy
DNS IP
DNS Eth
Phy
DNS
Bilkent’s NSP
network

 IP datagram forwarded from

campus network into comcast
network, routed (tables created
 IP datagram containing DNS by RIP, OSPF, IS-IS and/or
query forwarded via LAN BGP routing protocols) to DNS
switch from client to 1st hop server
router  demux’ed to DNS server
 DNS server replies to
client with IP address of
www.google.com
5: DataLink Layer 5-69
A day in the life… TCP connection carrying HTTP
HTTP
HTTP
SYNACK
SYN TCP
SYNACK
SYN IP
SYNACK
SYN Eth
Phy

 to send HTTP request,

client first opens TCP
socket to web server
 TCP SYN segment (step 1
SYNACK
SYN TCP in 3-way handshake) inter-
SYNACK
SYN IP domain routed to web
SYNACK
SYN Eth server
Phy  web server responds with
TCP SYNACK (step 2 in 3-
web server way handshake)
 TCP connection established!

5: DataLink Layer 5-70

A day in the life… HTTP request/reply
 web page finally (!!!)
displayed
HTTP
HTTP HTTP
HTTP
HTTP TCP
HTTP
HTTP IP
HTTP
HTTP Eth
Phy

 HTTP request sent into

TCP socket
 IP datagram containing
HTTP HTTP
TCP
HTTP request routed to
www.google.com
HTTP
HTTP IP
HTTP Eth  web server responds with
Phy HTTP reply (containing
web page)
web server  IP datgram containing
HTTP reply routed back to
client
5: DataLink Layer 5-71