Faculty:

BCS5O2 – COMPUTER NETWORKS Prof. Ashok Herur

[email protected]
Mobile: 91641 01399
 Module 2

Data Link Layer (DLL or L2)

Computer Networks - Prof Ashok Herur 2

Main topics in Module 2

Data link layer:

• Functions of DLL
• Framing & Synchronization
• Error detection and correction methods
• Elementary data link protocols, Sliding window protocols.

Medium access control sub-layer:

• Multiple access protocols.
• MAC in common LANs (Ethernet, Token Ring, Wireless LAN)

Computer Networks - Prof Ashok Herur 3

Functions of the Data Link Layer

• The data link layer is responsible for moving frames from one node to the
next one.
• Framing & Synchronization

• Error control over a link

• Error Detection and Correction codes

• Flow control over the links

Computer Networks - Prof Ashok Herur 4

Functions of the Data Link Layer

Computer Networks - Prof Ashok Herur 5

Framing

• It receives the “Packet” from the Network layer and adds a Header and a
Trailer to form a “Frame”

Computer Networks - Prof Ashok Herur 6

Framing

• A typical frame consists of the following overheads:

• Synchronization bits (typically alternating 1’s and 0’s) right at the beginning;
• Special Patterns to indicate the start and end of the frame;
• Redundant bits to help the receiver to deal with transmission errors;
• Physical (MAC or Medium Access Control) addresses of the two devices
across the link.

Computer Networks - Prof Ashok Herur 7

Framing - Synchronization

• The receiving device should be in perfect sync with the sending device.

• This means it should know the start of a bit duration and the exact
duration of a bit.

• If not perfectly synchronised, they may lose track, due to a small

difference (drift) in the clock rates of the transmitting and receiving
nodes.

• This will lead to bit-gain or bit-loss and even complete loss of the
message.

Computer Networks - Prof Ashok Herur 8

Framing – Start and End of the frame

• Special bit patterns, called Flags, are used to indicate the beginning and the
end of each frame.
• The popular HDLC (High-level Data Link Control) protocol used the same
pattern of 01111110 as the Start and End flags.
• One problem here is that the Flag pattern may occur in the data since it is a
bit-oriented protocol (non-text data like picture, audio, video, etc)
• This problem can be overcome by “bit-stuffing”.
• Sender scans the data and inserts an extra ‘0’ after 5 consecutive 1’s.
• Receiver removes the 0 after 5 consecutive 1’s.

Computer Networks - Prof Ashok Herur 9

Framing – Start and End of the frame

• Another protocol that is used is the Point-to-Point protocol (PPP).

• This also uses the same pattern of 01111110 as the Start and End flags.
• Unlike HDLC, this protocol supports both byte-oriented and bit-oriented
messages.
• It defines the:
• frame format of the data to be transmitted.
• procedure of establishing link between two points and exchange of data.
• method of encapsulation of network layer data in the frame.
• authentication rules of the communicating devices.

Computer Networks - Prof Ashok Herur 10

Framing

• Framing also involves computing the checksum (parity) bits and putting
them in the frame, to enable the receiver node to check for transmission
errors.
• It also includes the Physical (MAC) addresses of the sending node and the
receiving node.
• The MAC address of a network device is a permanent address that is
hardwired during its manufacture.
• A MAC address comprises of six groups of two hexadecimal digits (48
bits), separated by hyphens or colons.
• Example - 00:0A:89:5B:F0:11
• The IP address is used by routers to bring the packet to the right network
while the MAC address is used to deliver the frame to the device within the
network.
Computer Networks - Prof Ashok Herur 11
Error detection and correction codes

Computer Networks - Prof Ashok Herur 12

Error control

• When error detection codes are used, the receiver can only find out that
something is wrong but will not be able to say which bit(s) are incorrect.
• The only way out, is to ask the transmitter to retransmit the frame.
• Example codes: Character Parity, CRC
• When error correction codes are used, the receiver can pin-point the
erroneous bits and hence correct them.
• However, the amount of redundancy (overhead) here is much more than
in error detection codes.
• Example codes: Hamming code, Convolutional code.
• Error correction codes are also called Forward Error Correction (FEC) codes.

Computer Networks - Prof Ashok Herur 13

Error Detection codes versus Correction codes?

• When the error rate is very low (optical fibre), error detection codes are
better since the number of retransmissions will be very few.

• However, in real-time applications (process control, and even speech /

video too), we have to use error correction codes (if at all we wish to
incorporate error control).
• Do we need error control in audio / video?

• Transmission errors can be isolated (once in a while) or can occur in

bursts; The latter type are almost impossible to correct (though they can
be detected).

Computer Networks - Prof Ashok Herur 14

Character Parity coding

• An Even- or an Odd-parity scheme can be chosen.

• The number of 1’s after adding the encoding (Parity) bit should be
Even/Odd.
• The code of ‘A’ : 0100 0001 will be
• 0100 0001 0, in case of Even-parity scheme
• 0100 0001 1, in case of Odd-parity scheme

• Receiver checks for the required parity.

• Overhead is 12.5% (one extra bit for 8 message bits).
• Used for asynchronous transmission of characters (typically at a low
bit-rate)
Computer Networks - Prof Ashok Herur 15
Cyclic Redundancy Check (CRC) code

• These are powerful error detection codes, used for synchronous

transmission of large frames at high bit-rates.
• At high bit rates, the same impulsive noise will affect many bits and
hence the length of a burst error will be more.
• The Checksum is computed from the message using a Generator
Polynomial, and is then appended to the message.
• The receiver does the same computation and verifies the Checksum.
• For example, CRC-32 will generate 32 checksum bits, and can detect the
error even when 32 bits of a frame are corrupted.

Computer Networks - Prof Ashok Herur 16

CRC coding method – At the transmitter

• The coefficients of the Generator Polynomial (GP) are used to form a GP

string.
• Let the degree of the GP be “N”
• Then, N zeroes are appended to the message.
• The appended message is then divided by the GP string, using Modulo-2
arithmetic.
• No concept of Borrow during subtraction. Just XOR operation is done.
• The N-bit remainder is the Checksum. This is then appended to the original
message and is then transmitted.
• The receiver does the same computation and verifies the Checksum.
Computer Networks - Prof Ashok Herur 17
 CRC code – Example 1

• Compute the Checksum for the

message, 1101011111, using a
Generator Polynomial, x4 + x +
1

Computer Networks - Prof Ashok Herur 18

CRC coding method – At the receiver

• The receiver divides the received bit sting by the same GP string used by the
transmitter.
• If the remainder is Zero, it means that there are no errors. Else, it means
that the frame has some errors.
• Note that this method can detect the presence of up to N errors.

Computer Networks - Prof Ashok Herur 19

CRC code – Example 2

• Example:
1. Compute the Checksum for the message, 10010, using a Generator
Polynomial, x3 + x + 1
2. Show how the receiver verifies the Checksum when there is no error.
3. Show how the receiver detects the presence of a 2-bit error.

Computer Networks - Prof Ashok Herur 20

Hamming code

• Hamming code is an error correction code, that can detect and correct
single-bit errors.
• Hamming code uses a block parity mechanism; The data is divided into
blocks, and parity is added to the block.
• The number of parity bits added to Hamming code is given by the formula
2p ≥ d + p + 1, where p is the number of parity bits and d is the number of
data bits.
• For example, if you wanted to transmit 8 data bits, the formula would be 24
≥ 8 + 4 + 1. So 4 parity bits are required, for a total of 12 bits.
• This code is called Hamming (12, 8)

Computer Networks - Prof Ashok Herur 21

Hamming code

• These parity bits occupy the 2N bit positions, and are calculated by using
standard equations. The other bit places are filled by the bits representing
the character.
• Its transmission efficiency (code rate = k/n) goes up as the block size
increases.
• In Hamming(12, 8), the code rate (data bits / total) is only 0.67, while
Hamming(255, 247) is 0.969.
• However, as the block size increases, the chance of multiple errors in
the block also increases, and hence it will not be effective.
• It has a larger overhead compared to Character parity method.
• It needs 4 parity bits for each character (50% overhead) while Character
parity method required just one parity bit (12.5% overhead).
Computer Networks - Prof Ashok Herur 22
Example: Hamming (11, 7)

• The 4 equations for the computation of parity bits are:

• P1 + D3 + D5 + D7 + D9 + D11 = 0 (Even parity)
• P2 + D3 + D6 + D7 + D10 + D11 = 0
• P4 + D5 + D6 + D7 = 0
• P8 + D9 + D10 + D11 = 0
• The receiver computes the value of the 4 expressions (LHS only)
• If all of them give a value of ‘0’, it implies that there are no errors.
• If not, take the value of the 4 expressions and write them down in the
reverse order, and convert it to its decimal equivalent (let’s say, it’s E)
• The erroneous bit is the one in the Eth bit position.

Computer Networks - Prof Ashok Herur 23

Example: Hamming (11, 7)

1. Compute the Hamming parity bits for the message, 1001101.

2. Show how the receiver verifies that there is no error.
3. Show how the receiver detects and corrects the error in the 10th bit
position (of the received bit stream).

Solution:
• Since the message is 7 bits long, we need 4 parity bits, as per the formula:
2p ≥ d + p + 1.
• These parity bits will occupy the 1st, 2nd, 4th and 8th bit positions.

Computer Networks - Prof Ashok Herur 24

Example: Hamming (11, 7)

Computer Networks - Prof Ashok Herur 25

Example: Hamming (11, 7)

• The transmitted code is: 10011100101

• The received code (10th bit in error): 11011100101

• P1 + D3 + D5 + D7 + D9 + D11 =0
• P2 + D3 + D6 + D7 + D10 + D11 =1
• P4 + D5 + D6 + D7 =0
• P8 + D9 + D10 + D11 =1
• So, the number 1010 2 = 10 (decimal) indicates that bit number 10 is
incorrect. Hence, it is flipped from the received value of ‘1’ to ‘0’.

Computer Networks - Prof Ashok Herur 26

Convolutional code

• Convolutional code is an error correction code, that can detect and correct
multiple bit errors in a synchronous transmission frame.
• The computed parity bits depend not only on the current block but on the
past few blocks as well.
• It is quite complex and requires a lot of computation.
• It is also called as Viterbi coding.

Computer Networks - Prof Ashok Herur 27

Sliding window protocols

Computer Networks - Prof Ashok Herur 28

ARQ protocols

• When an error occurs, there is a retransmission request when error

detection codes are used.
• Common ARQ (Automatic Retransmission reQuest) protocols:
• Stop-and-wait protocol
• Go-back-N protocol
• Selective Repeat protocol
• The last two protocols are classified as “Sliding window protocols”

Computer Networks - Prof Ashok Herur 29

Stop-and-wait protocol

• Transmitter transmits a frame and waits for the acknowledgement (ACK),

which is sent by Receiver only if it receives the frame without any errors.
• It does not transmit the next frame till it receives the ACK for the
previous one.
• If ACK is not received within a specified time, the Transmitter times-out and
retransmits the frame.
• Results in a duplicate frame if ACK frame was lost.
• Need sequence numbers (0,1) to deal with this situation.
• Main drawback is the huge waiting time, and consequently the very low rate
of transmission (low utilization of bandwidth)

Computer Networks - Prof Ashok Herur 30

Stop-and-wait protocol

Computer Networks - Prof Ashok Herur 31

Sliding window protocols

• In these protocols, the sender is allowed to send up to N frames without

waiting for the acknowledgement of the earlier frames, where ‘N’ is the
window size.
• The sender maintains a set of sequence (frame) numbers corresponding to
frames sent but as yet not acknowledged. These frames are said to fall
within the ‘sending window’
• Similarly, the receiver maintains a ‘receiving window’ corresponding to
frames that it is permitted to accept.
• The two windows need not be of the same size.

Computer Networks - Prof Ashok Herur 32

Go-back-N protocol

• Sender’s window size = N, Receiver’s window size = 1

• The sender can send N frames without waiting for the acknowledgement,
but the receiver will only receive the one it is expecting to receive next.
• Subsequent frames, received after a faulty one, will be discarded till the
retransmitted frame is received.
• Whenever a frame is faulty, it has to be therefore retransmitted along with
all the subsequent frames that were sent earlier.
• No need to buffer frames at the receiver.

Computer Networks - Prof Ashok Herur 33

Go-back-N protocol (Nt = 4)

Computer Networks - Prof Ashok Herur 34

Selective Repeat Request protocol

• Sender’s window and the Receiver’s window are of the same size = N
• The sender can send N frames without waiting for the acknowledgement,
while the receiver can receive up to N frames.
• Whenever a frame is faulty, only the faulty frame has to be retransmitted.
• The receiver buffers the correct frames, while waiting for the retransmission
of the faulty frame.
• It then puts the frames in the correct sequence before passing them on to
the Network layer.

Computer Networks - Prof Ashok Herur 35

Selective Repeat Request protocol (N = 8)

Computer Networks - Prof Ashok Herur 36

Medium Access Control (MAC) protocols

Computer Networks - Prof Ashok Herur 37

Channel allocation problem

• The Channel allocation problem is about how to allocate a single broadcast

channel among many competing users.
• The network channel may be a single cable or optical fibre connecting
multiple nodes, or a portion of the wireless spectrum.
• Channel allocation algorithms allocate the wired channels and bandwidths
to the users, who may be base stations, access points or terminal equipment.
• The channel allocation can be Static or Dynamic.

Computer Networks - Prof Ashok Herur 38

Static Channel allocation

• In static channel allocation scheme, a fixed portion of the bandwidth is

allotted to each user.
• For N competing users, the bandwidth is divided into N channels, and each
portion is assigned to one user.
• In this allocation scheme, there is no interference between the users since
each user is assigned a fixed channel.
• However, it is not suitable in case of a large number of users (varying with
time) or when the traffic is bursty.
• It results in an inefficient use of the available bandwidth.

Computer Networks - Prof Ashok Herur 39

Dynamic Channel allocation

• In dynamic channel allocation scheme, users get to use the channels as and
when they are needed.
• The allocation protocols consider a number of parameters to minimize the
transmission interference from other users.
• This allocation scheme optimises bandwidth usage and results in faster
transmissions.
• Dynamic channel allocation is further divided into centralised (cell-phone
networks) and distributed allocation (LAN).

Computer Networks - Prof Ashok Herur 40

MAC protocols – Pure ALOHA

• The ALOHA protocol was developed at the University of Hawaii in the early
1970s for packet radio networks.
• However, it can be used in any situation where multiple devices share a
common communication channel.
• This protocol allows devices to transmit data at any time, without a set
schedule. There is no coordination between devices.
• When multiple devices attempt to transmit data at around the same time, it
can result in a collision. A collision is assumed when no ACK is received.
• In this case, each device will simply wait a random amount of time before
attempting to transmit again.

Computer Networks - Prof Ashok Herur 41

MAC protocols – Pure ALOHA

Computer Networks - Prof Ashok Herur 42

MAC protocols – Pure ALOHA

Computer Networks - Prof Ashok Herur 43

MAC protocols – Slotted ALOHA

• Here, the channel time is divided into time slots, and stations are only
authorized to transmit at the beginning of a time-slot.
• The duration of a time slot is exactly equal to the packet transmission
time.
• As a result, wasted time due to collisions can be reduced to one packet time
or the susceptible period can be half of that in Pure ALOHA.
• If the frame is received successfully, the receiver sends an acknowledgment.
• If the acknowledgment is not received within a time-out period, the
sender assumes that the frame was not received and retransmits the
frame in the next time slot.

Computer Networks - Prof Ashok Herur 44

 Throughput of Pure ALOHA and Slotted ALOHA

G = the offered load (or the number of packets being transmitted in one packet
time i.e. in one time-slot). In other words, it is a measure of the number of
nodes attempting to transmit in the time of a given time slot.
Computer Networks - Prof Ashok Herur 45
Carrier Sense Multiple Access (CSMA) protocols

• The maximum throughput in Slotted ALOHA is just 37%.

• It’s not surprising since the stations are transmitting at will, without
knowing what the other stations are doing.
• In LANs, it is often possible for stations to detect what other stations are
doing, and thus adapt their behaviour accordingly.
• Protocols in which stations “listen” for a “carrier” (transmission signal) and
act accordingly are called Carrier Sense protocols.
• In short, they don’t transmit when they know that others are transmitting,
thereby reducing the collisions, and improving the throughput.

Computer Networks - Prof Ashok Herur 46

Carrier Sense Multiple Access (CSMA) protocols

• Common CSMA protocols:

• 1-persistent protocol.
• Non-persistent protocol.
• p-persistent protocol.
• CSMA / CD (Collision Detection) protocol.
• Collision-free protocols:
• Bit-map protocol.
• Token passing protocol.

Computer Networks - Prof Ashok Herur 47

1-persistent protocol

• When the station is ready to send the frames, it will sense the channel.
• If the channel found to be busy, the station will wait for it to be idle.
• If the channel is found to be idle, the station transmits the frame
immediately.
• It is called “persistent” because it continuously tracks a busy channel till it
becomes idle.
• It is called “1-persistent” because after finding a channel to be idle, it surely
transmits it’s frame immediately (with a probability of 1, i.e. 100%)

Computer Networks - Prof Ashok Herur 48

1-persistent protocol

Computer Networks - Prof Ashok Herur 49

1-persistent protocol

• The problem with this method is that there is a high chance of collisions:
• When two or more stations find the channel in an idle state and they
transmit their frames at the same time.
• When the frames of a station are being transferred, then all the other
stations will sense that and won’t transmit the data; But the moment
they sense that the transmission has ended, they will all send their
frames at the same time.
• Another drawback is that when a collision occurs, the concerned stations
have to wait for a random time for the channel to be idle, and start all over
again.

Computer Networks - Prof Ashok Herur 50

Non-persistent protocol

• As in 1-persistent CSMA, the station that has frames to send will sense for
the channel.
• If the channel is found to be idle, the station transmits the frame
immediately.
• However, if the channel found to be busy, the station will wait for a
random time before sensing it again.
• In this method, the station does not continuously sense for the channel to
get idle.
• This reduces the chances of collision, but also reduces the efficiency of the
network when the load is light.

Computer Networks - Prof Ashok Herur 51

Non-persistent protocol

Computer Networks - Prof Ashok Herur 52

p-persistent protocol

• p-persistent CSMA is used when a channel has time-slots and the duration of a
time-slot is equal to or greater than the maximum propagation delay time for that
channel.

• When the station is ready to send the frames, it will sense the channel.
• If the channel found to be busy, the station will wait for the next time-slot.
• But if the channel is found to be idle, the station transmits the frame
immediately with a probability p.
• The station then waits for the beginning of the next time-slot with a probability
q = (1-p).
• If the next time-slot is also found idle, the station transmits or waits again with
the probabilities p and q.
• This process repeats until either the frame gets transmitted or another station
starts transmitting.
Computer Networks - Prof Ashok Herur 53
Throughput of all CSMA protocols

Computer Networks - Prof Ashok Herur 54

CSMA / CD protocol

Computer Networks - Prof Ashok Herur 55

CSMA / CD

• CSMA with CD (Collision Detection) is the most popular MAC protocol used
on the classic Ethernet LAN.

• Whenever a station has a frame to transmit, it will sense for a carrier

(signal) on the BUS (common transmission medium).

• It will not transmit if a carrier is found.

• If two stations transmit at around the same time, it will result in a collision.

• In that case, both the stations should suspend their transmissions, and try
after a random time later.

Prof. Ashok Herur 56

Ethernet LAN

Prof. Ashok Herur 57

CSMA / CD protocol

Prof. Ashok Herur 58

Binary Exponential Back-off algorithm

• Ethernet uses Binary exponential Back-off Algorithm to determine the

random waiting time in the event of a collision

• Delay time is (K * 51.2) us, where K is determined as follows:

▶1st collision: choose K randomly from {0,1}
▶2nd collision: choose K from {0,1,2,3}
▶ time collision: choose from {0,..,2 – 1}
■ Note max value for k = 1023, after 10 repeated collisions
▶give up after several tries (usually 16)
■ Report transmit error to host

Prof. Ashok Herur 59

Collision detection

• The NIC of a station will compare the signal on the Bus with what it had
actually transmitted.
• If they do not match, it means that there is a collision.
• Collision detection is a comparison done in the analog domain.

• Note that once the signal representing the first bit of a frame has reached
both the ends of the Bus, there will be no collisions since all the stations will
be able to sense the carrier.

Prof. Ashok Herur 60

10 Base 5 Ethernet LAN

• 10 Mbps bit-rate
• Base-band signalling is used (just pulses, no modulation)
• 5 x 100 = 500 metre is the maximum length of the Bus without a repeater
• Maximum of four repeaters allowed, implying a maximum overall length
of 2.5 kms for the Bus
• Also called “Thick Ethernet”, because of the thick coax used as the Bus.

Prof. Ashok Herur 61

Repeater
• It can be used to extend the length of the Bus in a LAN

Prof. Ashok Herur 62

 IEEE 802.3

• IEEE 802.3 is the standard defined for Ethernet LANs.

• Frame format (field length in Bytes):

7 1 2 or 6 2 or 6 2 0-1500 0-46 4

SFD Length of
Start of Frame Destination Address Source Address Data / Data Pad Frame Check Sequence
Preamble Delimiter Type of data

Prof. Ashok Herur 63

Performance of Ethernet LAN

• Performance can be measured by the “throughput”

• The throughput will be higher if the time lost due to collisions is lower
• Performance goes down with increasing load (many stations trying to
transmit at a given time)
• Performance is better if, on an average, the frame length is longer
• Limitations:
• There is no upper bound (limit) on the waiting time (for a transmission
to be successful).
• There is no Priority scheme; All stations are deemed to be “equal”

Prof. Ashok Herur 64

Bridge and Switch

• There are times when it is necessary to break up a large LAN into smaller
segments to improve the network performance.
• A Bridge is used to interconnect two LAN segments, while a Switch
interconnects 2 or more LAN segments.
• A Switch is just a multi-port bridge.
• The function of the bridge is to make intelligent decisions about whether
or not to pass signals between the two segments of a network.

Prof. Ashok Herur 65

Bridge and Switch

• When a bridge receives a frame on the network, the destination MAC

address is looked up in the bridge look-up table to determine whether to
Filter, Flood, or Forward the frame onto the other segment.
• Bridges and Switches automatically ‘learn’ about the topology and update
their look-up tables.
• If placed strategically, bridges and switches can greatly improve network
performance.

Prof. Ashok Herur 66

Hub
• Hubs are actually multiport repeaters.
• A repeater has just 2 ports, while a hub has 4 to 24 ports.

Prof. Ashok Herur 67

 10BaseT based Ethernet

Prof. Ashok Herur 68

Collision-free protocols

• Bit-map protocol.
• Token passing protocol.

Computer Networks - Prof Ashok Herur 69

Bit-map protocol

• Here, each contention period consists of exactly N slots. If any station has
to send a frame, then it transmits a ‘1’ bit in it’s corresponding slot.
• For example, if station 2 has a frame to send, it transmits a 1 bit in the
2nd slot.
• In this way, each station has complete knowledge of which station wishes
to transmit.
• There will never be any collisions because everyone agrees on who goes
next.
• Protocols like this in which the desire to transmit is broadcasted, for the
actual transmission, are called Reservation Protocols.

Computer Networks - Prof Ashok Herur 70

Bit-map protocol

Computer Networks - Prof Ashok Herur 71

Token Ring LAN

• Token Ring protocol is used on a Token Ring LAN.

• Unlike the Ethernet, this LAN provides:
• An upper bound on the waiting time (before a frame is successfully
transmitted);
• A priority scheme.

• Stations on a Token ring LAN are organized in a ring topology, with the data
being transmitted sequentially from one station to the next.

• The message sent by any station goes round the ring and is finally drained
off the ring by the transmitting station itself.

Prof. Ashok Herur 72

Token Ring LAN

• A control ‘token’ frame circulates around the ring to control access of the
ring.
• A station would ‘seize’ the token, send its message frame and release the
token.
• As the message frame circles around the ring, each station will observe the
destination address in the frame to decide whether to make a copy.
• IEEE 802.5 defines the standard for this type of LANs

Prof. Ashok Herur 73

Topology and transmission

Prof. Ashok Herur 74

Token Ring protocol

• After the message frame circles the ring and returns to its origin, the
originating station removes it from the ring and transmits a new token for
another station to use.
• A station cannot hold on to a token for more than the set Token-holding
time (default – 20 ms).

Prof. Ashok Herur 75

Access control byte

P P P T M R R R

PPP – Priority bits

RRR – Reservation bits

T – Token bit

M – Monitor bit

Prof. Ashok Herur 76

Token Ring protocol

• A token can be seized only if the message frame priority is equal or higher
than the token priority (PPP)
• A reservation to use the next token can be made, when a token or a
message is passing by, by writing the message priority in the RRR field
• Reservations can be overwritten by subsequent stations if the priority of
their message is higher than the current value in RRR.

Prof. Ashok Herur 77

Performance of Token Ring LAN

• There is no loss of throughput with increasing load, due to lack of collisions.

• Even when the load is light (few stations trying to transmit), a station has to
wait its turn for the token to arrive for its use.
• There is an upper bound (limit) on the waiting time.
• Priority and Reservation schemes can be implemented.

Prof. Ashok Herur 78

 Wireless LAN

Prof. Ashok Herur 79

Wireless LAN

• It is a LAN where data communication occurs through space, using

electromagnetic waves.
• A wireless LAN is implemented as an extension to, or as an alternative for,
a wired LAN.
• Benefits:
• Mobility of end user devices
• Fast and flexible installation
• No wires, no obstacles, no topology constraints
• Scalability

Prof. Ashok Herur 80

WLAN architecture

• Here clients, such as laptops and smart phones, connect to networks like
the company intranet or the Internet.

• Each client is associated with an Access Point (AP) that is in turn

connected to the other network.

• The client sends and receives its packets via the AP.

• The client has access to the network resources (server, shared printer, etc)
as well as to other clients.

Prof. Ashok Herur 81

Multiple Access Points

Printer

Access
Points

Prof. Ashok Herur 82

IEEE 802.11 standards

• IEEE 802.11 is the standard designed for wireless LANs

• It covers the Physical layer and the MAC sub-layer of the Data link layer.
• The first standard, published in 1997, was labeled 802.11
• Many newer versions for enhanced features and higher speeds were
rolled out subsequently as 802.11a, 802.11b, 802.11g and 802.11n

Prof. Ashok Herur 83

Why not use CSMA / CD?

• The CSMA/CD protocol used on wired Ethernet LANs cannot be used in

Wireless LANs, for two reasons:
• Carrier sensing is not effective because the transmission ranges of different
stations is different.
- This gives rise to problems like Hidden terminal problem.
- In a wired Ethernet, all stations can hear each other.
• Collision detection is almost impossible because the strength of the signal
received from another station can be just a millionth of the transmitted
signal strength.

Prof. Ashok Herur 84

Hidden terminal problem

If A “senses the
channel, it will not
hear C’s
transmission and will
falsely conclude that
it can begin a
transmission to B.

Prof. Ashok Herur 85

CSMA / Collision Avoidance

Prof. Ashok Herur 86

 CSMA / Collision Avoidance

• 802.11 tries to avoid collisions by using a protocol called CSMA/Collision

Avoidance (CSMA/CA)
• ‘A’ starts off by sending a short RTS (Request to Send) frame to ‘B’.
• This frame also contains the length of the data frame that is to follow.
• If ‘B’ is free, it responds with a CTS (Clear To Send) frame. This also
contains the length of the data frame, as copied from the RTS.
• Upon receipt of CTS, ‘A starts the actual data transmission.

Prof. Ashok Herur 87

 CSMA / Collision Avoidance: Typical issues

• Any station hearing the RTS is close to ‘A’, and will remain silent long
enough for the CTS to be transmitted back to A.
• Any station hearing the CTS is close to ‘B’, and will remain silent during the
upcoming transmission, whose length was mentioned in the CTS frame.
• ‘C’ is within the range of ‘A’ but not within the range of ‘B’; Therefore, it
hears the RTS but not the CTS.
• In contrast ‘D’ is within the range of ‘B’ but not within the range of ‘A’;
Therefore, it does not hear the RTS but hears the CTS.
• A station like ‘E” will hear both, and will not interfere.

Prof. Ashok Herur 88

 CSMA / Collision Avoidance: Typical issues

• Despite all these precautions, collisions can still occur.

• For example, B and C could both send RTS frames to A at the same time.
• These will collide and be lost.
• Both will therefore not receive a CTS, will wait for a random amount of
time and try again later.

Prof. Ashok Herur 89

Computer Networks - Prof Ashok Herur 90