Computer Networks, Fifth Edition by Andrew Tanenbaum and David Wetherall, Pearson Education-Prentice Hall, 2011

The Data Link Layer

Chapter 3
Computer Networks, Fifth Edition by Andrew Tanenbaum and David Wetherall, Pearson Education-Prentice Hall, 2011

Data Link Layer Design Issues
Network layer services
Framing
Error control
Flow control

Data Link Layer
Algorithms for achieving:
Reliable, +
Efficient,
communication of a whole units frames (as opposed to bits
Physical Layer) between two machines.
Two machines are connected by a communication channel
that acts conceptually like a wire (e.g., telephone line, coaxial
cable, or wireless channel).
Essential property of a channel that makes it wire-like
connection is that the bits are delivered in exactly the same
order in which they are sent.
Data Link Layer
For ideal channel (no distortion, unlimited bandwidth and no
delay) the job of data link layer would be trivial.
However, limited bandwidth, distortions and delay makes this
job very difficult.
Data Link Layer Design Issues
Physical layer delivers bits of information to and from data link
layer. The functions of Data Link Layer are:
1. Providing a well-defined service interface to the network
layer.
2. Dealing with transmission errors.
3. Regulating the flow of data so that slow receivers are not
swamped by fast senders.
Data Link layer
Takes the packets from Physical layer, and
Encapsulates them into frames
Data Link Layer Design Issues
Each frame has a
frame header a field for holding the packet, and
frame trailer.
Frame Management is what Data Link Layer does.

See figure in the next slide:
Computer Networks, Fifth Edition by Andrew Tanenbaum and David Wetherall, Pearson Education-Prentice Hall, 2011

Packets and Frames
Relationship between packets and frames.
Services Provided to the Network Layer
Principal Service Function of the data link layer is to transfer
the data from the network layer on the source machine to the
network layer on the destination machine.
Process in the network layer that hands some bits to the data
link layer for transmission.
Job of data link layer is to transmit the bits to the destination
machine so they can be handed over to the network layer there
(see figure in the next slide).
Computer Networks, Fifth Edition by Andrew Tanenbaum and David Wetherall, Pearson Education-Prentice Hall, 2011

Network Layer Services
(a) Virtual communication. (b) Actual communication.
Computer Networks, Fifth Edition by Andrew Tanenbaum and David Wetherall, Pearson Education-Prentice Hall, 2011

Possible Services Offered
1. Unacknowledged connectionless service.
2. Acknowledged connectionless service.
3. Acknowledged connection-oriented service.
Unacknowledged Connectionless
Service
It consists of having the source machine send independent
frames to the destination machine without having the
destination machine acknowledge them.
Example: Ethernet, Voice over IP, etc. in all the
communication channel were real time operation is more
important that quality of transmission.
Acknowledged Connectionless Service
Each frame send by the Data Link layer is acknowledged and
the sender knows if a specific frame has been received or lost.
Typically the protocol uses a specific time period that if has
passed without getting acknowledgment it will re-send the
frame.
This service is useful for commutation when an unreliable
channel is being utilized (e.g., 802.11 WiFi).

Network layer does not know frame size of the packets and
other restriction of the data link layer. Hence it becomes
necessary for data link layer to have some mechanism to
optimize the transmission.
Acknowledged Connection Oriented
Service
Source and Destination establish a connection first.
Each frame sent is numbered
Data link layer guarantees that each frame sent is indeed
received.
It guarantees that each frame is received only once and that all
frames are received in the correct order.
Examples:
Satellite channel communication,
Long-distance telephone communication, etc.

Acknowledged Connection Oriented
Service
Three distinct phases:
1. Connection is established by having both side initialize
variables and counters needed to keep track of which frames
have been received and which ones have not.
2. One or more frames are transmitted.
3. Finally, the connection is released freeing up the variables,
buffers, and other resources used to maintain the connection.

Framing
To provide service to the network layer the data link layer must
use the service provided to it by physical layer.
Stream of data bits provided to data link layer is not
guaranteed to be without errors.
Errors could be:
Number of received bits does not match number of transmitted
bits (deletion or insertion)
Bit Value
It is up to data link layer to correct the errors if necessary.
Framing
Transmission of the data link layer starts with breaking up the
bit stream
into discrete frames
Computation of a checksum for each frame, and
Include the checksum into the frame before it is transmitted.
Receiver computes its checksum error for a receiving frame
and if it is different from the checksum that is being
transmitted will have to deal with the error.

Framing is more difficult than one could think!
Computer Networks, Fifth Edition by Andrew Tanenbaum and David Wetherall, Pearson Education-Prentice Hall, 2011

Framing Methods
1. Byte count.
2. Flag bytes with byte stuffing.
3. Flag bits with bit stuffing.
4. Physical layer coding violations.
Byte Count Framing Method
It uses a field in the header to specify the number of bytes in
the frame.
Once the header information is being received it will be used
to determine end of the frame.
See figure in the next slide:
Trouble with this algorithm is that when the count is incorrectly
received the destination will get out of synch with
transmission.
Destination may be able to detect that the frame is in error but
it does not have a means (in this algorithm) how to correct it.
Computer Networks, Fifth Edition by Andrew Tanenbaum and David Wetherall, Pearson Education-Prentice Hall, 2011

Framing (1)
A byte stream. (a) Without errors. (b) With one error.
Flag Bytes with Byte Staffing Framing
Method
This methods gets around the boundary detection of the frame
by having each appended by the frame start and frame end
special bytes.
If they are the same (beginning and ending byte in the frame)
they are called flag byte.
In the next slide figure this byte is shown as FLAG.
If the actual data contains a byte that is identical to the FLAG
byte (e.g., picture, data stream, etc.) the convention that can
be used is to have escape character inserted just before the
FLAG character.
Computer Networks, Fifth Edition by Andrew Tanenbaum and David Wetherall, Pearson Education-Prentice Hall, 2011

Framing (2)
a) A frame delimited by flag bytes.
b) Four examples of byte sequences before and after byte stuffing.
Flag Bits with Bit Stuffing Framing
Method
This methods achieves the same thing as Byte Stuffing
method by using Bits (1) instead of Bytes (8 Bits).
It was developed for High-level Data Link Control (HDLC)
protocol.
Each frames begins and ends with a special bit patter:
01111110 or 0x7E <- Flag Byte
Whenever the senders data link layer encounters five
consecutive 1s in the data it automatically stuffs a 0 bit into the
outgoing bit stream.
USB uses bit stuffing.
Computer Networks, Fifth Edition by Andrew Tanenbaum and David Wetherall, Pearson Education-Prentice Hall, 2011

Framing (3)
Bit stuffing. (a) The original data. (b) The data as they appear on
the line. (c) The data as they are stored in the receivers memory after
destuffing.
Framing
Many data link protocols use a combination of presented
methods for safety. For example in Ethernet and 802.11 each
frame begin with a well-defined pattern called a preamble.
Preamble is typically 72 bits long.
It is then followed by a length fileld.
Error Control
After solving the marking of the frame with start and end the
data link layer has to handle eventual errors in transmission or
detection.
Ensuring that all frames are delivered to the network layer at
the destination and in proper order.
Unacknowledged connectionless service: it is OK for the
sender to output frames regardless of its reception.
Reliable connection-oriented service: it is NOT OK.
Error Control
Reliable connection-oriented service usually will provide a
sender with some feedback about what is happening at the
other end of the line.
Receiver Sends Back Special Control Frames.
If the Sender Receives positive Acknowledgment it will know
that the frame has arrived safely.
Timer and Frame Sequence Number for the Sender is
Necessary to handle the case when there is no response
(positive or negative) from the Receiver .
Flow Control
Important Design issue for the cases when the sender is
running on a fast powerful computer and receiver is running
on a slow low-end machine.
Two approaches:
1. Feedback-based flow control
2. Rate-based flow control
Feedback-based Flow Control
Receiver sends back information to the sender giving it
permission to send more data, or
Telling sender how receiver is doing.
Rate-based Flow Control
Built in mechanism that limits the rate at which sender may
transmit data, without the need for feedback from the receiver.
Error Detection and Correction
Two basic strategies to deal with errors:
1. Include enough redundant information to
enable the receiver to deduce what the
transmitted data must have been.

Error correcting codes.
2. Include only enough redundancy to allow
the receiver to deduce that an error has
occurred (but not which error).

Error detecting codes.
Error Detection and Correction
Error codes are examined in Link Layer because
this is the first place that we have run up against the
problem of reliability transmitting groups of bits.
Codes are reused because reliability is an overall
concern.
The error correcting code are also seen in the
physical layer for noise channels.
Commonly they are used in link, network and
transport layer.
Error Detection and Correction
Error codes have been developed after long
fundamental research conducted in mathematics.
Many protocol standards get codes from the large
field in mathematics.
Computer Networks, Fifth Edition by Andrew Tanenbaum and David Wetherall, Pearson Education-Prentice Hall, 2011

Error Detection & Correction Code (1)
1. Hamming codes.
2. Binary convolutional codes.
3. Reed-Solomon codes.
4. Low-Density Parity Check codes.
Error Detection & Correction Code
All the codes presented in previous slide add redundancy to
the information that is being sent.
A frame consists of
mdata bits (message) and
r redundant bits (check).
Block code - the r check bits are computed solely as function
of the m data bits with which they are associated.

Systemic code the m data bits are send directly along with
the check bits.
Linear code the r check bits are computed as a linear
function of the m data bits.
Error Detection & Correction Code
n total length of a block (i.e., n = m + r)
(n, m) code
n bit codeword containing n bits.
m/n code rate (range for noisy channel and
close to 1 for high-quality channel).

Error Detection & Correction Code
Example
Transmitted: 10001001
Received: 10110001
XOR operation gives number of bits that are different.
XOR: 00111000

Number of bit positions in which two codewords
differ is called Hamming Distance. It shows that two
codes are d distance apart, and it will require d
errors to convert one into the other.
Error Detection & Correction Code
All 2
m
possible data messages are legal, but due to
the way the check bits are computers not all 2
n

possible code words are used.
Only small fraction of 2
m
/2
n
=1/2
r
are possible will be
legal codewords.
The error-detecting and error-correcting codes of
the block code depend on this Hamming distance.
To reliably detect d error, one would need a
distance d+1 code.
To correct d error, one would need a distance 2d+1
code.
Error Detection & Correction Code
All 2
m
possible data messages are legal, but due to
the way the check bits are computers not all 2
n

possible code words are used.
Only small fraction of 2
m
/2
n
=1/2
r
are possible will be
legal codewords.
The error-detecting and error-correcting codes of
the block code depend on this Hamming distance.
To reliably detect d error, one would need a
distance d+1 code.
To correct d error, one would need a distance 2d+1
code.
Error Detection & Correction Code
Example:
4 valid codes:
0000000000
0000011111
1111100000
1111111111
Minimal Distance of this code is 5 => can correct double
errors and it can detect quadruple errors.
0000000111 => single or double bit error. Hence the
receiving end must assume the original transmission was
0000011111.
0000000000 => had triple error that was received as
0000000111 it would be detected in error.
Error Detection & Correction Code
One cannot perform double errors and at the same time detect
quadruple errors.
Error correction requires evaluation of each candidate
codeword which may be time consuming search.
Through design this search time can be minimized.
In theory if n = m + r, this requirement becomes:
(m + r + 1) 2
r

Hamming Code
Codeword: b1 b2 b3 b4 .
Check bits: The bits that are powers of 2
(p1, p2, p4, p8, p16, ).
The rest of bits (m3, m5, m6, m7, m9, )
are filled with m data bits.
Example of the Hamming code with m = 7
data bits and r =4 check bits is given in the
next slide.
The Hamming Code
Consider a message having four data bits (D) which is to be transmitted as a
7-bit codeword by adding three error control bits. This would be called a (7,4)
code. The three bits to be added are three EVEN Parity bits (P), where the
parity of each is computed on different subsets of the message bits as shown
below.

7 6 5 4 3 2 1
D

D

D

P
D

P P 7-BIT CODEWORD
D - D - D - P (EVEN PARITY)
D D - - D P - (EVEN PARITY)
D D D P - - - (EVEN PARITY)
Hamming Code
Why Those Bits? - The three parity bits (1,2,4) are related to
the data bits (3,5,6,7) as shown at right. In this diagram, each
overlapping circle corresponds to one parity bit and defines
the four bits contributing to that parity computation. For
example, data bit 3 contributes to parity bits 1 and 2. Each
circle (parity bit) encompasses a total of four bits, and each
circle must have EVEN parity. Given four data bits, the three
parity bits can easily be chosen to ensure this condition.
It can be observed that changing any one bit numbered 1..7
uniquely affects the three parity bits. Changing bit 7 affects all
three parity bits, while an error in bit 6 affects only parity bits 2
and 4, and an error in a parity bit affects only that bit. The
location of any single bit error is determined directly upon
checking the three parity circles.

Hamming Code
Hamming Code
For example, the message 1101 would be
sent as 1100110, since:
7 6 5 4 3 2 1
1 1 0 0 1 1 0 7-BIT CODEWORD
1 - 0 - 1 - 0 (EVEN PARITY)
1 1 - - 1 1 - (EVEN PARITY)
1 1 0 0 - - - (EVEN PARITY)
Hamming Codes
When these seven bits are entered into the
parity circles, it can be confirmed that the
choice of these three parity bits ensures that
the parity within each circle is EVEN, as
shown here.
Hamming Code
It may now be observed that if an error occurs in any of the
seven bits, that error will affect different combinations of the
three parity bits depending on the bit position.

For example, suppose the above message 1100110 is sent
and a single bit error occurs such that the codeword 1110110
is received:

transmitted message received message
1 1 0 0 1 1 0 ------------> 1 1 1 0 1 1 0
BIT: 7 6 5 4 3 2 1 BIT: 7 6 5 4 3 2 1

The above error (in bit 5) can be corrected by examining which of
the three parity bits was affected by the bad bit:
Hamming Code
7 6 5 4 3 2 1
1 1 1 0 1 1 0 7-BIT CODEWORD
1 - 1 - 1 - 0 (EVEN PARITY) NOT! 1
1 1 - - 1 1 - (EVEN PARITY) OK! 0
1 1 1 0 - - - (EVEN PARITY) NOT! 1
Hamming Code
In fact, the bad parity bits labeled 101 point directly to the bad
bit since 101 binary equals 5. Examination of the 'parity
circles' confirms that any single bit error could be corrected in
this way.
The value of the Hamming code can be summarized:
1. Detection of 2 bit errors (assuming no correction is
attempted);
2. Correction of single bit errors;
3. Cost of 3 bits added to a 4-bit message.
The ability to correct single bit errors comes at a cost which is
less than sending the entire message twice. (Recall that
simply sending a message twice accomplishes no error
correction.)

Computer Networks, Fifth Edition by Andrew Tanenbaum and David Wetherall, Pearson Education-Prentice Hall, 2011

Error Detection Codes (2)
Example of an (11, 7) Hamming code
correcting a single-bit error.
Convolutional Codes
Not a block code
There is no natural message size or encoding boundary as in
a block code.
The output depends on the current and previous input bits.
Encoder has memory.
The number of previous bits on which the output depends is
called the constraint length of the code.
They are deployed as part of the
GSM mobile phone system
Satellite Communications, and
802.11 (see example in the previous slide).
Computer Networks, Fifth Edition by Andrew Tanenbaum and David Wetherall, Pearson Education-Prentice Hall, 2011

Error Detection Codes (3)
The NASA binary convolutional code used in 802.11.
Convolutional Encoders
Like any error-correcting code, a convolutional code works by
adding some structured redundant information to the user's
data and then correcting errors using this information.
A convolutional encoder is a linear system.
A binary convolutional encoder can be represented as a shift
register. The outputs of the encoder are modulo 2 sums of the
values in the certain register's cells. The input to the encoder
is either the unencoded sequence (for non-recursive codes) or
the unencoded sequence added with the values of some
register's cells (for recursive codes).
Convolutional codes can be systematic and non-systematic.
Systematic codes are those where an unencoded sequence is
a part of the output sequence. Systematic codes are almost
always recursive, conversely, non-recursive codes are almost
always non-systematic.

Convolutional Encoders
A combination of register's cells that forms one of the output
streams (or that is added with the input stream for recursive
codes) is defined by a polynomial. Let m be the maximum
degree of the polynomials constituting a code, then K=m+1 is
a constraint length of the code.

Figure 1. A standard NASA convolutional
encoder with polynomials (171,133).
Convolutional Encoders
For example, for the decoder on the Figure 1, the polynomials
are:
g
1
(z)=1+z+z
2
+z
3
+z
6

g
2
(z)=1+z
2
+z
3
+z
5
+z
6

A code rate is an inverse number of output polynomials.
For the sake of clarity, in this article we will restrict ourselves
to the codes with rate R=1/2. Decoding procedure for other
codes is similar.
Encoder polynomials are usually denoted in the octal notation.
For the above example, these designations are 1111001 =
171 and 1011011 = 133.
The constraint length of this code is 7.
An example of a recursive convolutional encoder is on the
Figure 2.

Example of the Convolutional Encoder


Figure 2. A recursive convolutional
encoder.
Trellis Diagram
A convolutional encoder is often seen as a finite state
machine. Each state corresponds to some value of the
encoder's register. Given the input bit value, from a certain
state the encoder can move to two other states. These state
transitions constitute a diagram which is called a trellis
diagram.
A trellis diagram for the code on the Figure 2 is depicted on
the Figure 3. A solid line corresponds to input 0, a dotted line
to input 1 (note that encoder states are designated in such a
way that the rightmost bit is the newest one).
Each path on the trellis diagram corresponds to a valid
sequence from the encoder's output. Conversely, any valid
sequence from the encoder's output can be represented as a
path on the trellis diagram. One of the possible paths is
denoted as red (as an example).

Trellis Diagram
Figure 3. A trellis diagram corresponding to
the encoder on the Figure 2.
Trellis Diagram
Note that each state transition on the diagram corresponds to
a pair of output bits. There are only two allowed transitions for
every state, so there are two allowed pairs of output bits, and
the two other pairs are forbidden. If an error occurs, it is very
likely that the receiver will get a set of forbidden pairs, which
don't constitute a path on the trellis diagram. So, the task of
the decoder is to find a path on the trellis diagram which is the
closest match to the received sequence.

Trellis Diagram
Let's define a free distance d
f
as a minimal Hamming distance
between two different allowed binary sequences (a Hamming
distance is defined as a number of differing bits).
A free distance is an important property of the convolutional
code. It influences a number of closely located errors the
decoder is able to correct.

Viterbi Algorithm
Viterbi algorithm reconstructs the maximum-likelihood path for
a given input sequence.


Computer Networks, Fifth Edition by Andrew Tanenbaum and David Wetherall, Pearson Education-Prentice Hall, 2011

Error-Detecting Codes (1)
Linear, systematic block codes
1. Parity.
2. Checksums.
3. Cyclic Redundancy Checks (CRCs).
Parity Bit Error Detection
Block Size (m) 1000 bits from the equation:


we need (r) 10 check bits (1011 < 1024).
1 Mbit of data would require 10 kbits.
To detect a block with a single bit of error, one parity bit would
suffice.
Once every 1000 blocks (bit error rate is 10
-6
) one extra block
would have to be re-transmitted.
+ +1 2

Parity Bit Error Detection
Problem with multiple bit errors in burst errors.
Considerer the data stream as a matrix of n bits by k bits.
Each (k) row is computed a parity.
Interleaving is used to compute parity in different order from
that that is being transmitted.
Compute parity for each n columns
Transmit the data as k rows.
The last row we will send the n parity bits.

Example in the next slide gives the case for n=7 and k=7.
Computer Networks, Fifth Edition by Andrew Tanenbaum and David Wetherall, Pearson Education-Prentice Hall, 2011

Parity Error-Detecting Codes (2)
Interleaving of parity bits to detect a burst error.
Parity Error-Detecting Codes
A burst of length n+1 will pass undetected.
The probability of that any of the n columns
will have the correct parity by accident is
0.5; the probability of a bad block being
accepted as a good one is 2
-n
.
Checksum Error-Detecting Codes
A group of parity bits is one example of a
checksum.
Stronger checksums are based on a running sum of
the data bits of the message.
The checksum is usually placed at the end of the
message complementary sum.
Errors can be detected by summing the entire
received codeword, both data bits and checksum
bits.
If result is zero no error has been detected.
Checksum Error-Detecting Codes
Example if checksum is the 16-bit Internet error
detection used as part of the IP protocol.
It is applied to 16-bit words.
It will detect an error for cases where parity
detection fails.
Checksum error would fail for:
Deletion or addition of zero data,
Swapping part of the message,
Messages splices in which parts of two packets are
put together.
Those are typical errors caused by faulty hardware.
Checksum Error-Detecting Codes
Fletchers checksum
Includes positional component- adding the product
of the data and its position to the running sum.
The stronger kind of error-detecting code is Cyclic
Redundancy Check (CRC) know as polynomial
code.
A k-bit frame is regarded as the coefficient list for a
polynomial with k terms ranging from x
k-1
to x
0
CRC Error-Detecting Codes
Example:
110001: 1x
5
+ 1x
4
+ 0x
3
+ 0x
2
+ 0x
1
+ 1x
0

Module 2 arithmetic
Addition and Subtraction are equivalent to exclusive
OR.
Long division is carried the same way except that
subtraction operation is again done module 2.
CRC Error-Detecting Codes
Protocol requires that sender to agree in advance
with the receiver on the generator polynomial,
G(x).
Both high- and low-order bits must be 1.
CRC is computed for a frame of length m- bits
corresponding that is longer than the G(x).
When the receiver gets a checksummed frame it
divides it by G(x). If there the result is not equal to
zero it means that there has been transmission error.
CRC Error-Detecting Codes
Algorithm:
1. Let r be the degree of G(x). Append r zero bits to
the low-order end of the frame so it now contains
m+r bits and corresponds to the polynomial x
r
M(x)
2. Divide the bit string corresponding to G(x) into the
bit string corresponding to x
r
M(x) using modulo 2
division.
3. Subtract the remanider (which is always r or fewer
bits) from the bit string corresponding to x
r
M(x)
using module 2 subtraction. The result is the
checksummed frame, T(x), to be transmitted.
CRC Error-Detecting Codes
Example:
Frame: 1101011111
Generator: x
4
+x+1
Computer Networks, Fifth Edition by Andrew Tanenbaum and David Wetherall, Pearson Education-Prentice Hall, 2011

Error-Detecting Codes (3)
Example calculation of the CRC
Computer Networks, Fifth Edition by Andrew Tanenbaum and David Wetherall, Pearson Education-Prentice Hall, 2011

CRC Error-Detecting Codes
Certain polynomials have become international
standards: IEEE 802
x
32
+ x
26
+ x
23
+ x
22
+ x
16
+ x
12
+ x
11
+ x
10
+ x
8
+ x
7
+ x
4
+ x
2
+ x
1
+ 1
Computer Networks, Fifth Edition by Andrew Tanenbaum and David Wetherall, Pearson Education-Prentice Hall, 2011

Elementary Data Link Protocols (1)
Utopian Simplex Protocol
Simplex Stop-and-Wait Protocol
Error-Free Channel
Simplex Stop-and-Wait Protocol
Noisy Channel
Computer Networks, Fifth Edition by Andrew Tanenbaum and David Wetherall, Pearson Education-Prentice Hall, 2011

Elementary Data Link Protocols (2)
Implementation of the physical, data link, and network layers.
Elementary Data Link Protocols
Assumptions:
1. A wants to send a long stream of data to machine B.
2. It uses reliable connection-oriented service.
3. It assumes to have infinite supply of data ready to
send,
4. It does not have to wait for data to be produced.
5. Machines A and B do not crash
6. Data link layer treads the data as packets of pure
data whose every bit is to be delivered to the
destination's network layer.
Elementary Data Link Protocols
Data Link layer job is to:
FRAMEING: encapsulate the data packets in a frame
by adding the header and trailer.
Elementary Data Link Protocols
Error Correction and Detection:
Control information is added to header, and
Checksum is added to trailer.
Frame is then transmitted to the other machine.
Elementary Data Link Protocols
Library procedures:
to_physical_layer
from_physical_layer
wait_for_event(&event)
event object will contain the information what has
happened.
In the receiving end:
Data is being received and the checksum is being computed.
If the error has occurred the data link layer is being
informed: event = chksum_err
If the data has arrived undamaged: event = frame_arrival.

Computer Networks, Fifth Edition by Andrew Tanenbaum and David Wetherall, Pearson Education-Prentice Hall, 2011

Elementary Data Link Protocols (3)
Some definitions needed in the protocols to follow. These
definitions are located in the file protocol.h.
. . .
Computer Networks, Fifth Edition by Andrew Tanenbaum and David Wetherall, Pearson Education-Prentice Hall, 2011

Elementary Data Link Protocols (4)
Some definitions needed in the protocols to follow. These
definitions are located in the file protocol.h.
. . .
Computer Networks, Fifth Edition by Andrew Tanenbaum and David Wetherall, Pearson Education-Prentice Hall, 2011

Elementary Data Link Protocols (5)
Some definitions needed in the protocols to follow. These
definitions are located in the file protocol.h.
Computer Networks, Fifth Edition by Andrew Tanenbaum and David Wetherall, Pearson Education-Prentice Hall, 2011

Utopian Simplex Protocol (1)
A utopian simplex protocol.
. . .
Computer Networks, Fifth Edition by Andrew Tanenbaum and David Wetherall, Pearson Education-Prentice Hall, 2011

Utopian Simplex Protocol (2)
A utopian simplex protocol.
Utopian Simplex Protocol
Unrealistic
It does not handle flow control
It does not handle error correction
Resembles connectionless service that relies
on higher layer to solve those problems.
Computer Networks, Fifth Edition by Andrew Tanenbaum and David Wetherall, Pearson Education-Prentice Hall, 2011

Simplex Stop-and-Wait Protocol
for a Noisy Channel (1)
A simplex stop-and-wait protocol.
. . .
Computer Networks, Fifth Edition by Andrew Tanenbaum and David Wetherall, Pearson Education-Prentice Hall, 2011

Simplex Stop-and-Wait Protocol
for a Noisy Channel (2)
A simplex stop-and-wait protocol.
Simplex Protocol
Solves the problem of flow control.
Blocking acknowledgment of the reception of
the frame partially achieves this.
If acknowledgment frame gets lost the
protocol will fail.
Error correction is not implemented.
One direction only.
Sliding Window Protocols
Full duplex data transmission
Running two instances of one of the previous
protocols,
Each uses separate data link for simplex data traffic in
different directions (forward channel and reverse
channel).
Better idea is to use the same channel in both
directions.
A and B frames are intermixed with acknowledgment
frames from A and B.
Kind field in the header of the incoming frame will be
used to tell the difference.

Sliding Window Protocols
In addition to being able to use duplex
communication, the further protocol improvement
would be to send the acknowledgment frame
together with the new data frame. This provides for
a free ride.
The technique of delaying the acknowledgment for
a bit to be added to the next packet is know as
piggybacking.
This technique add another complication.
If the data frame is available within a short time
period (less than msec) it will use piggybacking.
Otherwise, it will send separate acknowledgment.
Sliding Window
At any instant of time the sender maintains a set of
sequences numbers that correspond to frames that
are permitted to be send.
Those frames are said to fall within a sending
window.
Receiver also maintains a receiving window that
corresponds to the set of frames that it is permitted
to accept.
Sliding Window
The sequence numbers within the senders window
represent the frames that have been sent or can be
sent but are not yet been acknowledged.
Whenever the a new packet arrives from the
network layer, it is given the next highest sequence
number, and
The upper edge of the window is advanced by one.
When acknowledgement arrives the lower edge is
advance by one.
This is how the protocol maintains the list of
unacknowledged frames.
Sliding Window Protocols
A sliding window of size 1, with a 3-bit sequence number.
(a) Initially. (b) After the first frame has been sent.
Sliding Window Protocols
A sliding window of size 1, with a 3-bit sequence number
(c) After the first frame has been received. (d) After the first acknowledgement
has been received.
Computer Networks, Fifth Edition by Andrew Tanenbaum and David Wetherall, Pearson Education-Prentice Hall, 2011

Sliding Window Protocols (1)
A positive acknowledgement with retransmission protocol.
. . .
Computer Networks, Fifth Edition by Andrew Tanenbaum and David Wetherall, Pearson Education-Prentice Hall, 2011

Sliding Window Protocols (2)
A positive acknowledgement with retransmission protocol.
. . .
Computer Networks, Fifth Edition by Andrew Tanenbaum and David Wetherall, Pearson Education-Prentice Hall, 2011

Sliding Window Protocols (3)
A positive acknowledgement with retransmission protocol.
One-Bit Sliding Window Protocol
Window size = 1
Stop-and-wait
Sender transmits a frame and waits for its
acknowledgment before it sends the next one.
Figure of next slide depicts such a protocol.
Computer Networks, Fifth Edition by Andrew Tanenbaum and David Wetherall, Pearson Education-Prentice Hall, 2011

One-Bit Sliding Window Protocol (1)
A 1-bit sliding window protocol.
. . .
Computer Networks, Fifth Edition by Andrew Tanenbaum and David Wetherall, Pearson Education-Prentice Hall, 2011

One-Bit Sliding Window Protocol (2)
A 1-bit sliding window protocol.
. . .
Computer Networks, Fifth Edition by Andrew Tanenbaum and David Wetherall, Pearson Education-Prentice Hall, 2011

One-Bit Sliding Window Protocol (3)
A 1-bit sliding window protocol.
Computer Networks, Fifth Edition by Andrew Tanenbaum and David Wetherall, Pearson Education-Prentice Hall, 2011

One-Bit Sliding Window Protocol (4)
Two scenarios for protocol 4. (a) Normal case. (b) Abnormal
case. The notation is (seq, ack, packet number). An asterisk
indicates where a network layer accepts a packet
Protocol Using Go-Back-N
Up-to-know the protocols depicted so far are based
on the assumption that the time required for
transmission and reception of the acknowledgment
packets is negligible.
If this assumptions is not true then round-trip time
can have important implications in the efficiency of
the communication and the bandwidth utilization.
Protocol Using Go-Back-N
Example:
50 kbps satellite channel
500 msec round-trip propagation delay.
Lets use Protocol 4 (previous slides) to send a 1000 bit frames
via the satellite.
t = 0: the sender starts the first frame
t = 20 msec: the frame has been completely sent.
t = 270 msec: the frame has fully arrived at the satellite.
t = 520 msec: the acknowledgment has arrived at the sender.

Sender was blocked 500/520 or 96% of the time. In order to
send the packed the sender utilized 4% of the available
bandwidth.

Protocol Using Go-Back-N
The problem?
Consequence of the rule that requires a sender to
wait for an acknowledgment before sending another
frame.
Relaxing this condition will enable achieving
significantly better throughput.
Solution!
Allowing sender the transmit w frames before
blocking.
Acknowledgment will arrive for previous frames
before the window becomes full.
Protocol Using Go-Back-N
Must correctly establish what is the actual size of
w.
Number of frames to fit inside the channel.
Capacity of the channel is determined by
Bandwidth in bits/sec
One-way transit time, or
The bandwidth-delay product of the link: BD
The actual size should be set to:
w = 2BD+1
Protocol Using Go-Back-N
Example:
Link with the bandwidth of 50 kbps
One way transit time of 250 msec
Bandwidth delay product is
BD = 50 kbps x 250 msec = 12.5 kbits or 12.5 frames
of 1000 bits.
2BD+1 = 26 frames.
t = 0: sends a first frame and subsequent frames after
20 msec.
t = 520 msec: acknowledgment for the first frame is
received, while 26 frames were transmitted.
Thereafter, acknowledgments will arrive every 20
msec.
Protocol Using Go-Back-N
Just in time for the sender to continue transmitting.
25 or 26 unacknowledged frames will always be
outstanding.
Hence, the senders maximum window size is 26
frames.
For smaller window size, the utilization of the link
will be less than 100%.

1 + 2

Upper bound of the link utilization:
It does not allow for any frame processing time
Treats acknowledgment frame as having zero length.
Protocol Using Go-Back-N
From the equation:
Large bandwidth-delay product requires a large
window.
With stop-and-wait for which w = 1, of there is even
one frames worth of propagation delay, the
efficiency will be less than 50%.
Pipelining technique where the number of frames
(much greater than 1) are send immediately.
What will happen then when a certain frame is
received in error?!
Computer Networks, Fifth Edition by Andrew Tanenbaum and David Wetherall, Pearson Education-Prentice Hall, 2011

Protocol Using Go-Back-N (1)
Pipelining and error recovery. Effect of an error when
(a) receivers window size is 1
Computer Networks, Fifth Edition by Andrew Tanenbaum and David Wetherall, Pearson Education-Prentice Hall, 2011

Protocol Using Go-Back-N (2)
Pipelining and error recovery. Effect of an error when
(b) receivers window size is large.
Computer Networks, Fifth Edition by Andrew Tanenbaum and David Wetherall, Pearson Education-Prentice Hall, 2011

Protocol Using Go-Back-N (3)
A sliding window protocol using go-back-n.
. . .
Computer Networks, Fifth Edition by Andrew Tanenbaum and David Wetherall, Pearson Education-Prentice Hall, 2011

Protocol Using Go-Back-N (4)
A sliding window protocol using go-back-n.
. . .
Computer Networks, Fifth Edition by Andrew Tanenbaum and David Wetherall, Pearson Education-Prentice Hall, 2011

Protocol Using Go-Back-N (5)
A sliding window protocol using go-back-n.
. . .
Computer Networks, Fifth Edition by Andrew Tanenbaum and David Wetherall, Pearson Education-Prentice Hall, 2011

Protocol Using Go-Back-N (6)
A sliding window protocol using go-back-n.
. . .
Computer Networks, Fifth Edition by Andrew Tanenbaum and David Wetherall, Pearson Education-Prentice Hall, 2011

Protocol Using Go-Back-N (7)
A sliding window protocol using go-back-n.
. . .
Computer Networks, Fifth Edition by Andrew Tanenbaum and David Wetherall, Pearson Education-Prentice Hall, 2011

Protocol Using Go-Back-N (8)
A sliding window protocol using go-back-n.
. . .
Computer Networks, Fifth Edition by Andrew Tanenbaum and David Wetherall, Pearson Education-Prentice Hall, 2011

Protocol Using Go-Back-N (9)
A sliding window protocol using go-back-n.
Cumulative Acknowledgment
When an acknowledgment comes in for frame n,
frames n-1, n-2, are also automatically
acknowledged.
Because protocol 5 has multiple outstanding
frames, it logically needs multiple timers, one per
outstanding frame.
Each frame times out independently of all the other
ones.
The pending timeouts from a linked list, with each
node of the list containing the
Number of clock ticks until the timer expires,
Frame being timed,
A pointer to the next node
Computer Networks, Fifth Edition by Andrew Tanenbaum and David Wetherall, Pearson Education-Prentice Hall, 2011

Protocol Using Go-Back-N (10)
Simulation of multiple timers in software. (a) The queued
timeouts (b) The situation after the first timeout has expired.
Selective Repeat
The go-back-n protocol works well if errors are rare.
If the line is poor it wastes a lot of bandwidth on
retransmitting frames.

A Selective Repeat protocol allows the receiver to
accept and buffer the frames following a damaged
or lost one.
Selective Repeat
Sender and Receiver maintain a window of
outstanding and acceptable sequence numbers,
respectively.
Sender:
Window size starts a 0,
Grows to some predefined maximum,
Receiver:
Is always fixed in size
Equal to the predetermined maximum.
Selective Repeat
Sender and Receiver maintain a window of
outstanding and acceptable sequence numbers,
respectively.
Sender:
Window size starts a 0,
Grows to some predefined maximum,
Receiver:
Is always fixed in size
Equal to the predetermined maximum.
Has a buffer reserved for each sequence number
within its fixed window.
Associated with it is a bit (arrived) telling whether
the buffer is full or empty.
Selective Repeat
Whenever the frame arrives, the receiver does the
following:
Check the frame sequence number if it does fall
within its fixed window,
If this is the case and the sequence has not been
received in previous transmissions it will be stored.
It does so without regard if the frame contains the
next packet expected by the network layer.
This frame is going to be kept until all the lower-
numbered frames have already been delivered to the
network layer in the correct order.
Selective Repeat
Consequential receive introduces further constrains
on frame sequence numbers vs. the protocols that
accepted frames in order.

Computer Networks, Fifth Edition by Andrew Tanenbaum and David Wetherall, Pearson Education-Prentice Hall, 2011

Protocol Using Selective Repeat (4)
A sliding window protocol using selective repeat.
. . .
Computer Networks, Fifth Edition by Andrew Tanenbaum and David Wetherall, Pearson Education-Prentice Hall, 2011

Protocol Using Selective Repeat (1)
A sliding window protocol using selective repeat.
. . .
Computer Networks, Fifth Edition by Andrew Tanenbaum and David Wetherall, Pearson Education-Prentice Hall, 2011

Protocol Using Selective Repeat (2)
A sliding window protocol using selective repeat.
. . .
Computer Networks, Fifth Edition by Andrew Tanenbaum and David Wetherall, Pearson Education-Prentice Hall, 2011

Protocol Using Selective Repeat (3)
A sliding window protocol using selective repeat.
. . .
Computer Networks, Fifth Edition by Andrew Tanenbaum and David Wetherall, Pearson Education-Prentice Hall, 2011

Protocol Using Selective Repeat (5)
A sliding window protocol using selective repeat.
. . .
Computer Networks, Fifth Edition by Andrew Tanenbaum and David Wetherall, Pearson Education-Prentice Hall, 2011

Protocol Using Selective Repeat (6)
A sliding window protocol using selective repeat.
. . .
Computer Networks, Fifth Edition by Andrew Tanenbaum and David Wetherall, Pearson Education-Prentice Hall, 2011

Protocol Using Selective Repeat (7)
A sliding window protocol using selective repeat.
. . .
Computer Networks, Fifth Edition by Andrew Tanenbaum and David Wetherall, Pearson Education-Prentice Hall, 2011

Protocol Using Selective Repeat (8)
A sliding window protocol using selective repeat.
. . .
Computer Networks, Fifth Edition by Andrew Tanenbaum and David Wetherall, Pearson Education-Prentice Hall, 2011

Protocol Using Selective Repeat (9)
A sliding window protocol using selective repeat.
Selective Repeat
Example:
3-bit sequence number,
7 frames are permitted to be send by transmitter,
before it is required to wait (for acknowledgment).
Window size restriction:
Must be at most half the range of the sequence
numbers.
Window size for protocol 6 is (MAX_SEQ+1)/2.
Computer Networks, Fifth Edition by Andrew Tanenbaum and David Wetherall, Pearson Education-Prentice Hall, 2011

Protocol Using Selective Repeat (10)
a) Initial situation with a window of size7
b) After 7 frames sent and received but not acknowledged.
c) Initial situation with a window size of 4.
d) After 4 frames sent and received but not acknowledged.
Selective Repeat
If there is a lot of traffic in one direction and no
traffic in the other direction, the protocol will block
when the sender window reaches its maximum.
Axillary timer is started by start_ack_timer after an
in-sequence data frame arrives.
If no reverse traffic has presented itself for
piggybacking before the timer expires, a separate
acknowledgement will be send.
Interrupt due to this timer is called ack_timeout
event.
Computer Networks, Fifth Edition by Andrew Tanenbaum and David Wetherall, Pearson Education-Prentice Hall, 2011

Example Data Link Protocols
1. Packet over SONET
2. ADSL (Asymmetric Digital Subscriber Loop)
Computer Networks, Fifth Edition by Andrew Tanenbaum and David Wetherall, Pearson Education-Prentice Hall, 2011

Packet over SONET (1)
Packet over SONET. (a) A protocol stack. (b) Frame relationships
Computer Networks, Fifth Edition by Andrew Tanenbaum and David Wetherall, Pearson Education-Prentice Hall, 2011

Packet over SONET (2)
Point-to-Point Protocol (PPP) Features
1. Separate packets, error detection
2. Link Control Protocol
3. Network Control Protocol

Computer Networks, Fifth Edition by Andrew Tanenbaum and David Wetherall, Pearson Education-Prentice Hall, 2011

Packet over SONET (3)
The PPP full frame format for unnumbered mode operation
Computer Networks, Fifth Edition by Andrew Tanenbaum and David Wetherall, Pearson Education-Prentice Hall, 2011

Packet over SONET (4)
State diagram for bringing a PPP link up and down
Computer Networks, Fifth Edition by Andrew Tanenbaum and David Wetherall, Pearson Education-Prentice Hall, 2011

ADSL (Asymmetric Digital
Subscriber Loop) (1)
ADSL protocol stacks.
Computer Networks, Fifth Edition by Andrew Tanenbaum and David Wetherall, Pearson Education-Prentice Hall, 2011

ADSL (Asymmetric Digital
Subscriber Loop) (1)
AAL5 frame carrying PPP data
Computer Networks, Fifth Edition by Andrew Tanenbaum and David Wetherall, Pearson Education-Prentice Hall, 2011

End
Chapter 3