www.jntuworld.

com

JNTUWORLD

The Transport Layer

THE TRANSPORT LAYER

Transport Entity: The hardware and/or software which makes use of services provided by the
N/W layer, (with in the transport layer) is called transport entity.
Transport Service Provider: Layers 1 to 4 are called Transport Service Provider.
Transport Service User: The upper layers i.e., layers 5to 7 are called Transport Service User.
Transport Service Primitives: Which allow transport users (application programs) to access
the transport service.
TPDU (Transport Protocol Data Unit): Transmissions of message between 2 transport entities
are carried out by TPDU. The transport entity carries out the transport service primitives by
blocking the caller and sending a packet the service. Encapsulated in the payload of this packet is
a transport layer message for the servers transport entity.

The task of the transport layer is to provide reliable, cost-effective data transport from the source
machine to the destination machine, independent of physical network or networks currently in
use.

(a). The Transport Service:

(i). Services provided to the Upper layers

(ii). Quality Of Service

(iii). Transport Service Primitives

(i). Services provided to the Upper layers:

As n/w layer is part of communication subnet and is run by carrier, the users have no
control over the (subnet)n/w layer. Therefore, another layer is put on top of the n/w layer to
improve the QOS, which is the TRANSPORT LAYER. This layer makes use of the help of
the services provided by the n/w layer with the help of Transport Entity.
1

www.jntuworld.com

www.jntuworld.com

JNTUWORLD

The Transport Layer

If a transport entity is informed halfway through a long transmission that its n/w
connection has been abruptly terminated, with no indication of what has happened to the data
currently in transit, it can set up a new network connection to the remote transport entity. Using
this new n/w connection, it can send a query to its peer asking which data arrived and which did
not, and then pick up from where it left off. Lost packets and mangled data can be detected and
compensated by transport layer.

With transport layer, it is possible for application programs to be written using a standard
set of primitives and to have these programs work on a wide variety of networks, without having
to worry about dealing with different subnets interfaces and unreliable transmission. i.e. it forms
the major boundary between the Transport Service Provider (the upper 4 layers) and the
Transport Service User(the upper layer(s)).

(ii). Quality Of Service: The primary function of transport layer is to enhance the QOS provided by
the n/w layer. The transport user specifies all these QOS parameters where a connection is
requested. These are specified by the transport user when a connection is requested. Both the desired
and minimum acceptable values can be given. If these are unachievable, the transport layer informs
the caller that the connection attempt failed, (without even contacting the destination) along with
failure reasons. QOS can be characterized by a number of specific parameters such as:

www.jntuworld.com

www.jntuworld.com

JNTUWORLD

The Transport Layer

Connection establishment delay.

Connection establishment failure probability.

Throughput.

Transit delay.

Residual error ratio.

Protection.

Priority.

Resilience.

The Connection Establishment Delay is the amount of time elapsing between a

transport connection being requested and the configuration being received by the
transport service.

The Connection Establishment Failure Probability is the chance of a connection not

being established with in the maximum establishment delay time, due to some internal
problems.

The Throughput parameter measures the number of bytes of user data transferred per
second, measured over some time interval.

The Transit Delay measures the time between a message being sent by the transport user
or the source machine and it is being received by the transport user on the destination
machine.

The residual error ratio measures the number of lost or garbled messages as a fraction
of the total sent.

The protection parameter provides a way for the transport user to specify interest in
having the transport layer provide protection against unauthorized third parties (wire
tappers) reading or modifying the transmitted data.

The resilience parameter gives the probability of the transport layer itself spontaneously
terminating a connection due to internal problems or congestion.

The priority parameter provides a way for the transport user to indicate that some of its
connections are made important than others and in the event of congestion, to make sure
that high-priority connections get serviced before the low-priority ones.

www.jntuworld.com

www.jntuworld.com

JNTUWORLD

The Transport Layer

OPTION NEGOTIATION:
If the transport layer achieve the desired goal, but can achieve a lower, but still
acceptable rate, it then informs the remote machine about this and asks for connection
establishment. If the remote machine cannot handle the proposed value, but it can handle a value
above the minimum, it permits. If it cannot handle any value above the minimum, it rejects the
connection attempt. Finally, the originating transport user is informed of whether the connection
was established, the values of parameters agreed upon. This process is called Option
Negotiation.

(iii). Primitives for a simple Transport service:

Eg: Consider an application with a server and a number of remote clients.

1. The server executes a LISTEN primitive by calling a library procedure that makes a
system call to block the server until a client turns up.
2. when a client wants to talk to the server, it executes a CONNECT primitive, with
CONNECTION REQUEST TPDU sent to the server.
3. When it arrives, the TE unblocks the server and sends a CONNECTION ACCEPTED
TPDU back to the client.
4. When it arrives, the client is unblocked and the connection is established. Data can now
be exchanged using SEND and RECEIVE primitives.

www.jntuworld.com

www.jntuworld.com

JNTUWORLD

The Transport Layer

5. When a connection is no longer needed, it must be released to free up table space with in
the 2 transport entries, which is done with DISCONNECT primitive by sending
DISCONNECTION REQUEST TPDU. This disconnection can b done either by
asymmetric variant (connection is released, depending on other one) or by symmetric
variant (connection is released, independent of other one).

Primitives sent as TPDUs using encapsulation:

Berkely Sockets: These primitives are socket primitives used in Berkely UNIX for TCP.

www.jntuworld.com

www.jntuworld.com

JNTUWORLD

The Transport Layer

(b). Elements of Transport Protocols:

The transport service is implemented by a transport protocol between the 2 transport entities. The
elements are:

(i). Addressing

(ii). Connection establishment

(iii). Connection release

(iv). Flow control and buffering

(v). multiplexing

(vi). Crash recovery
(i). Addressing:
When an application process wishes to set up a connection to a remote application
process, it must specify which one to connect it to, i.e., TSAPs and NSAPs must be specified by
the transport layer and n/w layer respectively.

TSAP-----Transport Service Access Point

NSAP---- Network Service Access Point

(ii). Connection establishment:

With packet lifetimes bounded, it is possible to devise a fool proof way to establish
connections safely. Packet lifetime can be bounded to a known maximum using one of the
following techniques:

 Restricted subnet design

 Putting a hop counter in each packet
 Time stamping in each packet
Using a 3-way hand shake, a connection can be established. This establishment protocol doesnt
require both sides to begin sending with the same sequence number.

www.jntuworld.com

www.jntuworld.com

JNTUWORLD

Fig-(a)

The Transport Layer

Fig-(b)

Fig-(c)

Host-1 chooses a sequence

The first TPDU is a delayed

Host-2 gets a delayed

number, x and sends a

duplicate CONNECTION

CONNECTION REQUEST and

CONNECTION REQUEST

REQUEST from an old

replies to it. At this point, it is

TPDU containing it to host-2.

connection. This TPDU arrives

crucial to realize that host-2 has

host-2 replies with a

at host-2 without host-1s

proposed using y as the initial

CONNECTION ACCEPTED

knowledge. Host-2 reacts to this sequence number for host-2 to

TPDU acknowledging x and

TPDU by sending host-1 a

host-1 traffic, knowing that no

announcing its own initial

CONNECTION ACCEPTED

TPDUs containing sequence

sequence number y. finally,

TPDU, in effect asking for

number y or

host-1 acknowledges host-2s

verification that host-1 was

acknowledgements to y are

choice of an initial sequence

indeed trying to set up a new

still in existence. When the

number in the first data

connection. When host-1 rejects

second delayed TPDU arrives at

TPDU that it sends.

host-2s attempt to establish,

host-2, the fact that z has been

host-2 realizes that it was

acknowledged rather than y tells

tricked by a delayed duplicate

host-2 that this ,too, is an old

and abandons the connection.

duplicate.

www.jntuworld.com

www.jntuworld.com

JNTUWORLD

The Transport Layer

(iii). Connection release:

A connection is released using either asymmetric or symmetric variant. But, the
improved protocol for releasing a connection is a 3-way handshake protocol.

Fig-(a)

Fig-(b)

Fig-(c)

Fig-(d)

One of the user sends a

Initial process is

If the second

Same as in fig-( c)

DISCONNECTION REQUEST TPDU in

done in the same

DR is lost, the

except that all

order to initiate connection release. When

way as in fig-(a).

user initiating

repeated attempts to

it arrives, the recipient sends back a DR-

if the final ACK-

the

retransmit the DR

TPDU, too, and starts a timer. When this

TPDU is lost, the

disconnection

are assumed to be

DR arrives, the original sender sends back

situation is saved

will not receive

failed due to lost

an ACK-TPDU and releases the

by the timer.

the expected

TPDUs. After N

connection. Finally, when the ACK-TPDU

When the timer is

response, and

entries, the sender

arrives, the receiver also releases the

expired, the

will timeout and just gives up and

connection.

connection is

starts all over

releases the

released.

again.

connection.

www.jntuworld.com

www.jntuworld.com

JNTUWORLD

The Transport Layer

(iv). Flow control and Buffering:

Flow control is done by having a sliding window on each connection to keep a fast
transmitter from over running a slow receiver.
Buffering must be done by the sender, if the network service is unreliable. The sender
buffers all the TPDUs sent to the receiver. The buffer size varies for different TPDUs. They are:
 Chained Fixed-size Buffers
 Chained Variable-size Buffers
 One large Circular Buffer per Connection
(a). Chained Fixed-size Buffers:
If most TPDUs are nearly the same size, the buffers are organized as a pool of identical
size buffers, with one TPDU per buffer.

(b). Chained Variable-size Buffers:

This is an approach to the buffer-size problem. i.e., if there is wide variation in TPDU
size, from a few characters typed at a terminal to thousands of characters from file transfers,
some problems may occur:
 If the buffer size is chosen equal to the largest possible TPDU, space will be wasted when
ever a short TPDU arrives.
 If the buffer size is chosen less than the maximum TPDU size, multiple buffers will be
needed for long TPDUs.
To over come these problems, we employ variable-size buffers.

( c). One large Circular Buffer per Connection:

A single large circular buffer per connection is dedicated when all connections are
heavily loaded.
1. Source Buffering is used for low band width bursty traffic
2. Destination Buffering is used for high band width smooth traffic.
3. Dynamic Buffering is used if the traffic pattern changes randomly.

www.jntuworld.com

www.jntuworld.com

JNTUWORLD

The Transport Layer

(v). Multiplexing:
In networks that use virtual circuits with in the subnet, each open connection consumes some
table space in the routers for the entire duration of the connection. If buffers are dedicated to the
virtual circuit in each router as well, a user who left a terminal logged into a remote machine ,
there is need for multiplexing. There are 2 kinds of multiplexing:
1. Up-ward multiplexing: In the below figure, all the 4 distinct transport connections use the
same network connection to the remote host. When connect time forms the major component
of the carriers bill, it is up to the transport layer to group port connections according to their
destination and map each group onto the minimum number of port connections.

2. Down-ward multiplexing:

If too many transport connections are mapped onto the one network connection, the
performance will be poor.

If too few transport connections are mapped onto one network connection, the service
will be expensive.

The possible solution is to have the transport layer open multiple connections and distribute the
traffic among them on round-robin basis, as indicated in the below figure:

With k network connections open, the effective band width is increased by a factor of k.

10

www.jntuworld.com

www.jntuworld.com

JNTUWORLD

The Transport Layer

(vi). Crash recovery:

To recover from server crashes..
If there is a transmission between a client and a server and mean while if the server crashes, the
server will send a broadcast TPDU to all other hosts, announcing that it just crashed. Each client
can be in any of 2 states:

One TPDU outstanding, SI

No TPDUs outstanding, SO

THE INTERNET TRANPORT PROTOCOLS (TCP AND UDP):

The internet has 2 main protocols in the transport layer they are:
1. Connection-Oriented Protocol ( TCP )
2. Connection-Less Protocol

( UDP)

TCP ( TRANSMISSION CONTROL PROTOCOL )

It was specifically designed to provide a reliable end-to end byte stream over an unreliable
network. It was designed to adapt dynamically to properties of the inter network and to be
robust in the face of many kinds of failures.
Each machine supporting TCP has a TCP transport entity, which accepts user data
streams from local processes, breaks them up into pieces not exceeding 64kbytes and sends
each piece as a separate IP datagram. When these datagrams arrive at a machine, they are given
to TCP entity, which reconstructs the original byte streams. It is up to TCP to time out and
retransmits them as needed, also to reassemble datagrams into messages in proper sequence.

The different issues to be considered are:

a. The TCP Service Model
b. The TCP Protocol
c. The TCP Segment Header
d. The Connection Management
e. TCP Transmission Policy
f. TCP Congestion Control
g. TCP Timer Management.

11

www.jntuworld.com

www.jntuworld.com

JNTUWORLD

The Transport Layer

a). The TCP Service Model:

TCP service is obtained by having both the sender and receiver create end points called
SOCKETS

Each socket has a socket number(address)consisting of the IP address of the host, called a
PORT ( = TSAP )

To obtain TCP service a connection must be explicitly established between a socket on
the sending machine and and a socket on the receiving machine

All TCP connections are full duplex and point to point i.e., multicasting or broadcasting
is not supported.

A TCP connection is a byte stream, not a message stream i.e., the data is delivered as
chunks

Eg:

4 * 512 bytes of data is to be transmitted.

Urgent data is another feature of TCP in which the sending application puts some control
information in the data stream when an interactive user hits the DEL or CTRL-C key to break off
a remote computation that has already begun, and gives it to TCP along with URGENT flag.
This event causes TCP to stop accumulating data and transmit everything it has for that
connection immediately.

When the urgent data are received at the destination, the receiving application is
interrupted and reads the data stream to find the urgent data.

12

www.jntuworld.com

www.jntuworld.com

JNTUWORLD

The Transport Layer

Sockets:
A socket may be used for multiple connections at the same time. In other words, 2 or
more connections may terminate at same socket. Connections are identified by socket identifiers
at same socket. Connections are identified by socket identifiers at both ends. Some of the sockets
are listed below:

Ports: Port numbers below 256 are called Well- known ports and are reserved for standard
services.

Eg:
PORT-21

To establish a connection to a host to transfer a file using FTP

PORT-23

To establish a remote login session using TELNET

b). The TCP Protocol:

Every byte on a TCP connection has its own 32-bit sequence number, used both for
acknowledgement and window mechanism. The sending and receiving TCP entities exchange
data in the form of segments [20 byte header + optional part] followed by zero or more data
bytes] segments can be united or splitted, by the TCP s/w accordingly, based on 2 limits:

1. each segment must fit in 65,535 byte IP payload

2. Each segment must fit in MTU (Max. Transfer Unit) of a network.
13

www.jntuworld.com

www.jntuworld.com

JNTUWORLD

The Transport Layer

The basic protocol used by TCP entities is the SLIDING WINDOW

PROTOCOL when a sender transmits a segment, it also starts a timer. When the segment
arrives at destination, the receiving TCP entity sends back a segment having an
acknowledgement number equal to the next sequence number it expects to receive. If the sender
timer goes off before the acknowledgement is received, the sender transmits the segment again.

TCP at the design time must deal with different problems like:
 Since segments can be fragmented, it is possible that part of a transmitted segment arrives
and is acknowledged, while the rest is lost.
 Segments can also arrive out of order
 Segments can also be delayed so long in transit that the sender times out and retransmits
them
 If retransmitted segment takes other route than the original and reaches the receiver,
duplicates must be differentiated from the original ones.
 Segments may occasionally hit congested networks along its path.
c). The TCP Segment Header:

Every segment begins with a fixed-format 20 byte header followed by header options. Then
follows, up to 65,535-20-20=65,515 data bytes if any.

14

www.jntuworld.com

www.jntuworld.com

JNTUWORLD

The Transport Layer

Different fields of TCP header are:

 Source Port, Destination Port : Identify local end points of the connections
 Sequence number : Specifies the sequence number of the segment
 Acknowledgement Number : Specifies the next byte expected.
 TCP header length : Tells how many 32-bit words are contained in TCP header
 URG : It is set to 1 if URGENT pointer is in use, which indicates start of urgent data.
 ACK : It is set to 1 to indicate that the acknowledgement number is valid.
 PSH : Indicates pushed data
 RST : It is used to reset a connection that has become confused due to reject an invalid
segment or refuse an attempt to open a connection.
 FIN : Used to release a connection.
 SYN : Used to establish connections.
SYN ACK

Meaning

Piggyback acknowledgement field is not in use

Connection reply

d). TCP Connection management:

Connections are established in TCP using the 3-way handshake. To establish a

connection between a server and a client following actions take place at both sides:

SERVER

CLIENT

Passively waits for an incoming connection by Executes a CONNECT primitive specifying

executing LISTEN and ACCEPT primitives, the IP address and port to which it wants to
either specifying a specific source or nobody connect, the maximum TCP segment size it is
in particular

willing to accept and additionally some user

data.

15

www.jntuworld.com

www.jntuworld.com

JNTUWORLD

The Transport Layer

TCP connection establishment in normal case:

1). When a TCP segment arrive at
destination, the TCP entity
then ,checks to see if there is a
process that has done a
LISTEN on port given in
destination port field.
2). If not, it sends a reply with
RST bit-ON to reject the
connection other wise, the
process is given incoming
TCP segment, by which the
connection can be accepted or
rejected.
3). Acknowledgement is sent
back, if accepted.

TCP connection management finite state machine:

16

www.jntuworld.com

www.jntuworld.com

JNTUWORLD

The Transport Layer

TCP Connection management from servers point of view:

1. The server does a LISTEN and settles down to see who turns up.
2. When a SYN comes in, the server acknowledges it and goes to the SYNRCVD state
3. When the servers SYN is itself acknowledged the 3-way handshake is complete and server
goes to the ESTABLISHED state. Data transfer can now occur.
4. When the client has had enough, it does a close, which causes a FIN to arrive at the server
[dashed box marked passive close].
5. The server is then signaled.
6. When it too, does a CLOSE, a FIN is sent to the client.
7. When the clients acknowledgement shows up, the server releases the connection and deletes
the connection record.

e). TCP Transmission policy:

Window management in TCP:

17

www.jntuworld.com

www.jntuworld.com

JNTUWORLD

The Transport Layer

1. In the above example, the receiver has 4096-byte buffer.

2. If the sender transmits a 2048-byte segment that is correctly received, the receiver will
acknowledge the segment.
3.

Now the receiver will advertise a window of 2048 as it has only 2048 of buffer space, now.

4.

Now the sender transmits another 2048 bytes which are acknowledged, but the advertised
window is0.

5. The sender must stop until the application process on the receiving host has removed some
data from the buffer, at which time TCP can advertise a layer window.

SILLY WINDOW SYNDROME:

This is one of the problems that ruin the TCP performance, which occurs when data are
passed to the sending TCP entity in large blocks, but an interactive application on the receiving
side reads 1 byte at a time.

 Initially the TCP buffer on the receiving side is full and the sender knows this(win=0)
 Then the interactive application reads 1 character from tcp stream.
 Now, the receiving TCP sends a window update to the sender saying that it is all right to
send 1 byte.
 The sender obligates and sends 1 byte.
 The buffer is now full, and so the receiver acknowledges the 1 byte segment but sets
window to zero. This behavior can go on forever.
18

www.jntuworld.com

www.jntuworld.com

JNTUWORLD

The Transport Layer

f). TCP Congestion control:

TCP does to try to prevent the congestion from occurring in the first place in the following way:
When a connection is established, a suitable window size is chosen and the receiver
specifies a window based on its buffer size. If the sender sticks to this window size, problems
will not occur due to buffer overflow at the receiving end. But they may still occur due to
internal congestion with in the n/w. lets see this problem occurs.

Error!

In fig (a):

We see a thick pipe leading to a small- capacity receiver. As long as the sender

does not send more water than the bucket can contain, no water will be lost.

In fig (b): The limiting factor is not the bucket capacity, but the internal carrying capacity of the
n/w. if too much water comes in too fast, it will backup and some will be lost.

When a connection is established, the sender initializes the congestion window to the size
of the max segment in use our connection. It then sends one max segment .if this max segment is
acknowledged before the timer goes off, it adds one segment s worth of bytes to the congestion
window to make it two maximum size segments and sends 2 segments. As each of these segment
is acknowledged, the congestion window is increased by one max segment size. When the
congestion window is n segments, if all n are acknowledged on time, the congestion window
is increased by the byte count corresponding to n segments. The congestion window keeps
growing exponentially until either a time out occurs or the receivers window is reached.

19

www.jntuworld.com

www.jntuworld.com

JNTUWORLD

The Transport Layer

The internet congestion control algorithm uses a third parameter, the threshold in
addition to receiver and congestion windows.

Different congestion control algorithms used by TCP are:

 RTT variance Estimation.
 Exponential RTO back-off

Re-transmission Timer Management

 Karns Algorithm
 Slow Start
 Dynamic window sizing on congestion
 Fast Retransmit

Window Management

 Fast Recovery
 RTT Variance estimation:
Also known as Jacobsons algorithm. This enables a TCP entity to adapt to changes in
round-trip time. Mean-deviation is used as a variation measure and the value is used as
input in the calculation of RTO as.

MDEV(x)

E [ |x-E(x)| ]

Where E(x) is the expected value for x

AERR(K+1)
Where AERR(K+1)
ARTT(K)

RTT(K+1)-ARTT(K)

= the sample mean deviation measured at time k

= sample average round trip time for first k segments.

The complete algorithm proposed by JACOBSON is

Therefore

SRTT(K+1)

[(1-g)][SRTT(K)] + [G][RTT(K+1)]

SERR(K+1)

RTT(K+1) - SRTT(K)

SDEV(K+1)

[(1-h)][SDEV(K)] + [h][|SERR(K+1)|]

RTO(K+1)

SRTT(K+1) + [f][SDEV(K+1)]

Where SRTT is an exponentially smoothed estimate of RTT and the values of g, h, f are
given by g=1/8=0.125, h=1/4=0.25, f=2

20

www.jntuworld.com

www.jntuworld.com

JNTUWORLD

The Transport Layer

 Exponential RTO Back-off:

A TCP source increases its RTO each time a segment is retransmitted, when is
referred to as a BACKOFF process. After the first retransmission of a segment on each
effected connection, the TCP sources will all wait a longer time to clear the current
congestion. If a second retransmission is required, each TCP source will then wait an even
longer time before timing out for a 3rd retransmission, giving the internet an even longer
period to recover. A simple technique for implementing RTO back-off is to multiply the
RTO for a segment by a constant value for each retransmission:

RTO

q*RTO, which causes RTO to grow exponentially, with each retransmission this technique is
referred to as Binary Exponential Back-off where q=2

 Karns algorithm:
1. Do not use the measured RTT for a retransmitted segment to update SRTT and SDEV
2. Calculate back-off RTO using equation RTO=q*RTO when a retransmission occurs.
3. Use the back-off RTO value for sending segments until an acknowledgement arrives
for a segment that has not been transmitted.

 Slow Start:
JACOBSON recommended this procedure in which a congestion window is
defined which is measured in segments rather than octets. At any time, TCP transmission
is constrained by the following relationship.

Where

awnd =

MIN [ credit, cwnd ]

awnd =

allowed window in segments

cwnd =

congestion window in segments

credit =

the amount of unused credit granted in the most recent

acknowledgement in segments.

When a new connection is opened, the TCP entity initializes cwnd=1. Each time an
acknowledgement is received, the value of cwnd is increased by 1 up to some maximum value.
cwnd grows exponentially when the first ACK arrives, TCP opens cwnd=2 and can send 2
segments. When these 2 segments are acknowledged, TCP can send 4 segments and so on.

21

www.jntuworld.com

www.jntuworld.com

JNTUWORLD

The Transport Layer

 Fast retransmit:
If a TCP entity receives a segment out of order, it must immediately issue an
ACK for the last in order segment that was received. TCP will continue to repeat this
ACK with each incoming segment until the missing segment arrives to plug the hole in
its buffer. When the hole is plugged, TCP sends a cumulative ACK for all of the in order
segments received so for when a source TCP receives a duplicate ACK, it is made to wait
to receive 3 duplicate ACKs to same segment. Retransmission takes place, if 4 ACKs are
not received by sender.

 Fast recovery:
Retransmit the last segment, cut the cwnd in half and then proceed with the linear
increase of cwnd. This process can be explained as:

22

www.jntuworld.com

www.jntuworld.com

JNTUWORLD

The Transport Layer

1. When the third duplicate ACK arrives

a.

set

ssthresh

b.

retransmit the missing segment

c.

set

cwnd

cwnd/2

ssthresh+3

2. Each time an additional duplicate ACK arrives, increment cwnd by a 1 and

transmit a segment if possible
3. When a next ACK arrives that acknowledges new data, set cwnd = ssthresh

 Dynamic window sizing on congestion:

Jacobson proposed the use of slow start to begin with, followed by a linear growth
in cwnd. The rules are as follows, when a timeout occurs:
1. Set slow-start threshold equal to current congestion window, ssthresh = cwnd/2
2. Set

cwnd

= 1 and perform slow start process until cwnd = ssthresh

3. For

cwnd >=

ssthresh, increases cwnd by one for each round trip time

 Limited transmit:
This algorithm calls for a TCP sender to transmit a new segment the following 3
conditions are must.
1. 2 consecutive duplicate acknowledgements are received
2. The destination TCP entitys advertised allows the transmission of the segment
3. The amount of outstanding data, after sending the new segment, would remain less
than or equal to cwnd+2

 TCP timer management:

TCP uses 3 kinds of timers:
1. Retransmission timer
2. Persistence timer
3. Keep-Alive timer.

1.

Retransmission timer: When a segment is sent, a timer is started. If the segment is

acknowledged before the timer expires, the timer is stopped. If on the other hand, the timer goes

23

www.jntuworld.com

www.jntuworld.com

JNTUWORLD

The Transport Layer

off before the acknowledgement comes in, the segment is retransmitted and the timer is started
again. The algorithm that constantly adjusts the time-out interval, based on continuous
measurements of n/w performance was proposed by JACOBSON and works as follows:
 for each connection, TCP maintains a variable RTT, that is the best current estimate of the
round trip time to the destination inn question.
 When a segment is sent, a timer is started, both to see how long the acknowledgement takes
and to trigger a retransmission if it takes too long.
 If the acknowledgement gets back before the timer expires, TCP measures how long the
measurements took say M
 It then updates RTT according to the formula
RTT = RTT + ( 1-
)M
Where = a smoothing factor that determines how much weight is given to
the old value. Typically, =7/8
Retransmission timeout is calculated as
D

D + ( 1-
) | RTT-M |

another smoothed variable, Mean Deviation

RTT

expected acknowledgement value

observed acknowledgement value

Timeout

RTT+(4*D)

Where

2. Persistence timer:
It is designed to prevent the following deadlock:
 the receiver sends an acknowledgement with a window size of 0 telling the sender to wait
 later, the receiver updates the window, but the packet with the update is lost
 now both the sender and receiver are waiting for each other to do something
 when the persistence timer goes off, the sender transmits a probe to the receiver
 the response to the probe gives the window size
 if it is still zero, the persistence timer is set again and the cycle repeats
 if it is non zero, data can now be sent
3. Keep-Alive timer: When a connection has been idle for a long time, this timer may go off to
cause one side to check if other side is still there. If it fails to respond, the connection is
terminated.

24

www.jntuworld.com

www.jntuworld.com

JNTUWORLD

The Transport Layer

UDP ( USER DATAGRAM PROTOCOL )

It provides a connection-less service for application level procedures. UDP is present on the top
of IP. UDP adds a port addressing capability to IP. A UDP segment consists of an 8byte header,
followed by data. The header is as follows:

Source port, destination port: Identifies the end points with in the source and destination
machines.
UDP length: Includes 8-byte header and the data
UDP checksum: Includes the UDP header, the UDP data padded out to an even number o
Of bytes if need be. It is an optional field

The ATM AAL ( ATM Adaptation Layer) Layer protocols

AAL layer is present on top of ATM layer and is similar to transport layer, in some aspects. The
goal of AAL is to provide useful services to applications programs and to shield then from the
mechanics of chopping data up into cells at the source and reassembling them at the destination.
As ATM layer just outputs 53-byte cells consecutively, and no error control and flow control.
AAL was defined by ITU.
Original service classes supported by AAL:
A,B,C,D are the only classes that were of any use among all 8 district classes.

To handle these 4 classes of service, ITU defined 4 protocols respectively.

AAL - 1

AAL 2

AAL 3

AAL 4

25

www.jntuworld.com

www.jntuworld.com

JNTUWORLD

The Transport Layer

Later, as the technical requirements for classes C&D were similar, they were combined into
AAL - 3/4. AAL - 5 was also defined to meet further requirements

Structure of AAL:

The ATM adaptation layer is divided into 2 major parts:

 Convergence Sub layer
 SAR ( Segmentation And Reassembly ) Sub layer
Convergence layer:
The upper part of AAL is convergence sub layer. Its job is to provide the interface to the
application. It consists of a subpart that is common to all applications and an application specific
subpart. The functions of each of these parts are protocol dependent but can include message
framing and error detection. It is responsible for accepting bit streams or arbitrary length
messages from the applications and breaking them up into units of 44 to 48 bytes for trans at the
destination, this sub layer reassembles the cells into the original messages. Message boundaries
are preserved, if present.

SAR sub layer:

The lower part of AAL is SAR sub layer. It can add headers and trailers to the data units
given to it by the convergence sub layer to form cell payloads. These payloads are then given to
the ATM layer for transmission. At the destination, the SAR sub layer reassembles the cells into
messages. It sometimes handles error detection and multiplexing

26

www.jntuworld.com

www.jntuworld.com

JNTUWORLD

The Transport Layer

Operation:

From the above figure, when a message comes into the AAL from the application, the
convergence sub layer may give it a header and/or trailer. The message is then broken up into
44-to-48 byte units, which are passed to the SAR sub layer. The SAR sub layer may add its own
header and trailer to each piece and pass them down to the ATM layer for transmission as
independent cells.

AAL - 1:
 It is the protocol used for transmitting class-A traffic, that is real-time, constant bit rate,
connection-oriented traffic, such as uncompressed audio and video.
 Bits are fed in by the application at a constant rate and must be delivered at the far end, at
the same constant rate, with a minimum of delay, jitter, and overhead.
 The o/p is a stream of bits, with no message boundaries.
 Missing cells are reported to the application
 AAL-1 uses a Convergence & SAR sub layers, in which
Convergence sub layer: Detects lost and mis-inserted cells and also smoothes out incoming
traffic to provide delivery of cells at a constant rate. Finally, it breaks up the i/p messages into
46 or 47 byte units and are given to SAR for transmission. At the other end, it extracts these
and reconstructs the original input.

SAR sub layer: It has a protocol and 2 different formats are present for its cells.
27

www.jntuworld.com

www.jntuworld.com

JNTUWORLD

The Transport Layer

 Both formats begin with a 1-byte header containing a 3-bit cell sequence number, SN ( to
detect missing or mis-inserted cells).
 This field is followed by a 3-bit sequence number protection, SNP, over the sequence
number to allow correction of single errors and detection of double errors in the sequence
field. It uses CRC with polynomial x3+x+1
 An even parity bit covering the header byte further reduces the likelihood of a bad
sequence number sneaking in unnoticed
 The p cells are used when message boundaries must be preserved
 The pointer field is used to give offset of start of next message, and its range is 0 to 92

AAL - 2:
 This protocol has been designed where message boundaries must be preserved so that the
start of the next full frame can be recognized, even in the presence of lost cells or bad data
like compressed audio or video.

 It uses Convergence and SAR sub layers, in which

Convergence sub layer : it does not have a protocol and works same as in AAL-1
SAR sub layer : It has a protocol and its cell format is

It has a 1-byte header composing of SN(sequence no),which is used for numbering cells in
order to detect missing or mis-inserted cells and IT(information type) indicate that the cell is
the start, middle or end of the message.

The LI(length indicated)field tells how big the payload is, in bytes(<=45 bytes)

CRC field is a checksum over the entire cell, so errors can be detected

28

www.jntuworld.com

www.jntuworld.com

JNTUWORLD

The Transport Layer

AAL - 3/4:
 This protocol can be operate in 2 modes
1. stream mode, in which the boundaries are not preserved
2. message mode, in which the boundaries are preserved
 it supports multiplexing, which allows multiple sessions from a single host to travel along
the same virtual circuit ( Eg: remote logins ) and be separated at the destination
 It has both the convergence sub layer protocol and a SAR sub layer protocol ,in which
Convergence sub layer: messages come into this layer from application and are first padded
out to a multiple of 4 bytes then a header &trailer are attached.
CPI

Btag

BAsize

Payload ( 1 to 65,535bytes )

padding

Etag

CS header

Length 0 to 65535

CS trailer

CPI ( Common Part Indicator ): gives the message type and counting unit for the BA
size and length fields
B tag, E tag: used to frame messages and are incremented after each new message sent
by which lost or mis-inserted calls can be checked
BA size: used for buffer allocation i.e., tells the receiver to allocate how much bufferspace, for the message in advance of its arrival.
Length: gives the payload length again
SAR sub layer: the message from CS layer is received and chopped into 44-byte chunks
and inserted into payload of a cell whose format is

Error!

ST (Segment Type): used for message framing and indicates whether the cell begins a message,
is in the middle of a message, is the last cell of a message or is a small
(i.e., single cell) message.

29

www.jntuworld.com

www.jntuworld.com

JNTUWORLD

The Transport Layer

SN( Sequence Number ): used for detecting lost or mis-inserted cells.

MID( Multiplexing ID ): used to keep track of which cell belongs to which session (as CS layer
may buffer several messages belonging to different sessions and send
pieces of them in random order)All the pieces from messages
belonging to session i carry i in MID field so that they can be
correctly reassembled at destination.
LI+CRC: payload length + checksum of cell.

Disadvantages of AAL 3/4:

1. 2 layers of protocol overhead i.e., 8 bytes are added to every message; 4 bytes are added
to every cell.
2. It is a heavy weight mechanism for short messages.

AAL - 5:
Because of the complexity & Inefficiency generated by 2 layers of protocol in AAL3/4, coupled
with short checksum led to the invention of SEAL ( Simple Efficient Adaptation Layer )
protocol i.e., AAL-5.
 It offers unreliable service, with options to have cells with checksum errors either
discarded or passed to the application any way
 It offers reliable service, with few control to prevent over sums
 It supports 2 modes:

Stream mode

Message mode, in which an application can pass a datagram of length 1 to

65,535 bytes to AAL layer and have it delivered to destination.

 It has both a Convergence Sub layer Protocol and SAR Sub layer Protocol in which,
CS protocol: Receives message from the application and adds a trailer (8-byte) the
amount of padding ( < +47bytes )is chosen to make the entire message, including the
padding and trailer, be a multiple of 48 bytes. It doesnt have any header.
Payload(1to65,535bytes)

UU

Length

30

www.jntuworld.com

CRC

www.jntuworld.com

JNTUWORLD

The Transport Layer

UU(User- to-User): available for a CS-service specific sub layer for its own purpose
Eg: sequencing or multiplexing
Length: tells how long the true payload is, in bytes. A value of 0 is used to abort the
current message mid stream.
CRC: standard 32-bit checksum over the entire message
SAR sub layer : The message from CS, along with trailer is received and is breaked into
48-bytes units, each passed to ATM layer for transmission. PTI field is
set, if message boundaries are to be preserved.

Advantages of AAL-5:

Great efficiency.

It has no overhead in each cell.

Longer checksum calculates lost, mis-inserted or missing cells without using seq. numbers.

31

www.jntuworld.com