Transport Layer

A transport-layer protocol provides for logical communication between

applications processes of source machine and destination machine.

The hardware and/or software within the transport layer that does the work is
called the transport entity.

We will use the term segment for messages sent from transport entity to
transport entity. TCP, UDP and other Internet protocols use this term.

2
 A transport-layer protocol provides logical communication between processes
running on source and destination hosts, a network-layer protocol provides logical
communication between adjacent hosts. This distinction between these two layers is
subtle but important.

The transport layer provides logical

rather (lot of buffer or memory and
wide range of link types between
them) than physical communication
(short circuit no buffer) between
application processes.

3
Household Analogy
Consider two houses, one on the Dhaka and the other on the Khulna, with each house
being home to a dozen kids. The kids in the Dhaka household are cousins of the kids in
the Khulna household.

 The kids in the two households love to write to each other—each kid writes each cousin
every week, with each letter delivered by the traditional postal service in a separate
envelope. Thus, each household sends 144 letters (12×12 i.e. all to all relation) to the
other household every week.

Hasan Roshid

Dhaka Khulna

4
In each of the households there is one caretaker—Roshid in the Khulna house and
Hasan in the Dhaka house—responsible for mail collection and mail distribution.

Each week Roshid visits all brothers and sisters, collects the mail, and gives the mail to
a postal-service mail carrier, who makes daily visits to the house.

When letters arrive at the Khulna house, Roshid also has the job of distributing the
mail to her brothers and sisters. Hasan has a similar job on the Dhaka.

Hasan Roshid

Dhaka Khulna

5
In this example, the postal service provides logical communication between the two
houses (using intermediate post offices) —the postal service moves mail from house to
house, not from person to person.

On the other hand, Roshid and Hasan provide logical communication among the
cousins. Roshid and Hasan pick up postal mail from, and deliver mail to, their brothers
and sisters.

 application messages = letters in envelopes

 transport-layer protocol = Roshid and Hasan
 processes = cousins (who order Hansan/Roshid to send the letter)
 hosts (also called end systems) = houses
 network-layer protocol = postal service (including mail carriers), who deals with
several post offices hop-by-hop until reach the destination house.

6
On the sending side, the transport layer converts the application-layer
messages it receives from a sending application process into transport-
layer packets, known as transport-layer segments in Internet terminology.

This is done by (possibly) breaking the application messages into smaller

chunks and adding a transport-layer header to each chunk to create the
transport-layer segment.

The Internet has two main protocols in the transport layer, a connectionless
protocol and a connection-oriented one. In the following sections we will
study both of them. The connectionless protocol is UDP (User Datagram
Protocol). The connection-oriented protocol is TCP (Transmission Control
Protocol ).
7
 HTTP(which uses port number 80) and FTP(which
uses port number 21)

Two clients, using the same destination port number (80) to communicate
with the same Web server application. 8
Connection Establishment

 Establishing a connection sounds easy, but it is actually surprisingly

tricky. At first glance, it would seem sufficient for one transport entity
to just send a CONNECTION REQUEST TPDU to the destination and
wait for a CONNECTION ACCEPTED reply.

 The problem occurs when the network can lose, store, and duplicate
packets. This behavior causes serious complications.

9
The worst possible nightmare is as follows. A user establishes a
connection with a bank (online money transfer), sends messages telling
the bank to transfer a large amount of money to the account of a not-
entirely-trustworthy person, and then releases the connection.

Unfortunately, each packet in the scenario is duplicated and stored in the

subnet. After the connection has been released, all the packets pop out of
the subnet and arrive at the destination in order, asking the bank to
establish a new connection, transfer money (again), and release the
connection.

The bank has no way of telling that these are duplicates. It must assume
that this is a second, independent transaction, and transfers the money
again.
10
 Seq = x
Seq = y
Tomlinson (1975) introduced the three-way
handshake. The normal setup procedure when host 1
initiates is shown in fig. below. Host 1 chooses a
sequence number, x, and sends a CONNECTION
REQUEST TPDU containing it to host 2.

Host 2 replies with an ACK TPDU acknowledging x

and announcing its own initial sequence number, y.

Finally, host 1 acknowledges host 2's choice of an

initial sequence number in the first DATA TPDU that
it sends.

seq = x and ACK = y is sent only on first data segment but not on
the subsequent segments.
11
Now let us see how the three-way handshake works
in the presence of delayed duplicate control TPDUs.

In Fig. below, the first TPDU is a delayed duplicate

CONNECTION REQUEST from an old connection.
This TPDU arrives at host 2 without host 1's
knowledge. Host 2 reacts to this TPDU by sending
host 1 an ACK TPDU, in effect asking for verification
that host 1 was indeed trying to set up a new
connection.

When host 1 rejects host 2's attempt to establish a

connection, host 2 realizes that it was tricked by a
delayed duplicate and abandons the connection. In
this way, a delayed duplicate does no damage.

12
The worst case is when both a delayed
CONNECTION REQUEST (CR) and an ACK
are floating around in the subnet. This case is
shown in Fig. below.

As in the previous example, host 2 gets a

delayed CONNECTION REQUEST (CR) and
replies to it using y as the initial sequence number.
This CR will be rejected by Host-1.

When the second delayed TPDU:

DATA(seq = x, ACK = z) arrives at host 2, the
fact that z has been acknowledged rather than y
tells host 2 that this, is an old duplicate.

13
Connection Release
Releasing a connection is easier than establishing one. Nevertheless,
there are more pitfalls than one might expect. There are two styles of
terminating a connection: asymmetric release and symmetric release.

Asymmetric release is the way the telephone system works: when one
party hangs up, the connection is broken. Symmetric release treats the
connection as two separate unidirectional connections and requires each
one to be released separately.

14
 CR → Connection Request
 Asymmetric release is abrupt and
may result in data loss. Consider the
scenario of Fig. below. After the
connection is established, host 1
sends a TPDU that arrives properly
at host 2. Then host 1 sends another
TPDU. Unfortunately, host 2 issues a
DISCONNECT before the second
TPDU arrives. The result is that the
connection is released and data are
lost.

 Clearly, a more sophisticated release

protocol is needed to avoid data loss.
DR → Disconnection Request
15
One way is to use symmetric release, in which each
direction is released independently of the other one. Here, a
host can continue to receive data even after it has sent a
DISCONNECT TPDU.

Symmetric release does the job when each process has a

fixed amount of data to send and clearly knows when it has
sent it.

16
One can envision a protocol in which host 1 says: I am done. Are you done too? If host 2
responds: I am done too. Goodbye, the connection can be safely released. Unfortunately,
this protocol does not always work. There is a famous problem that illustrates this issue. It
is called the two-army problem.

Two-way handshaking experiences

two-army problem.

Suppose that the commander of blue army #1 sends a message reading: ‘‘I propose we attack at dawn on
March 29. How about it?’’ Now suppose that the message arrives, the commander of blue army #2 agrees,
and his reply gets safely back to blue army #1. Will the attack happen? Probably not, because commander #2
does not know if his reply got through. If it did not, blue army #1 will not attack, so it would be foolish for
him to charge into battle 17
We will next consider using a three-way
handshake for four scenarios of releasing
connection.
In Fig. (a), we see the normal case in which
one of the users sends a DR
(DISCONNECTION REQUEST) TPDU to
initiate the connection release.
When DR arrives, the recipient sends back a
DR TPDU, too, and starts a timer, just in case
its DR is lost.
When this DR arrives, the original sender
sends back an ACK TPDU and releases the
connection.
Finally, when the ACK TPDU arrives, the
receiver also releases the connection.

18
If the final ACK TPDU is lost, as shown in Fig. 6-14(b), the situation is saved by the timer.
When the timer expires, the connection is released anyway.

19
Now consider the case of the second DR being lost. The user initiating the disconnection
will not receive the expected response, will time out, and will start all over again. In Fig.
(c) we see how this works, assuming that the second time no TPDUs are lost and all
TPDUs are delivered correctly and on time.

20
Our last scenario, Fig. (d), is the same as Fig. (c) except that now we assume all the
repeated attempts to retransmit the DR also fail due to lost TPDUs. After N retries, the
sender just gives up and releases the connection. Meanwhile, the receiver times out and
also exits.

21
Multiplexing and Demultiplexing

22
 At the destination host, the transport layer receives
segments (packet or data segment) from the network Application layer
layer just below it. The transport layer has the
responsibility of delivering the data segments to the
appropriate application process running in the host. Transport layer

 Suppose you are sitting in front of your computer, and Network layer
you are downloading Web pages while running one
FTP session and two Telnet sessions. You therefore
have four network application processes running—two
Telnet processes, one FTP process, and one HTTP
process.

23
 When the transport layer in your computer receives data from the network
layer below, it needs to direct the received data to one of these four
processes.

 The transport layer in the receiving host does not actually deliver data
directly to a process, but instead to an intermediary socket.

Application layer
Each process is associated
Transport layer with a socket.

Network layer

24
 This job of delivering the data in a transport-layer segment to the correct socket is
called demultiplexing (S/P converter)

 The job of gathering data chunks at the source host from different sockets,
(encapsulating each data chunk with header information to create segments) and
passing the segments to the network layer is called multiplexing (P/S converter).

 Note that the transport layer in the middle host in fig. below must demultiplex
segments arriving from the network layer to either process P1 or P2.

25
Two clients, using the same destination port number (80) is communicating
with the same Web server application.
26
TCP Header
Source Port (16 bits): Source TCP user. Example values are Telnet 23;A complete
list is maintained at http://www.iana.org/
assignments/port-numbers.
Destination Port (16 bits): Destination TCP user.
Sequence Number (32 bits): Sequence number of the first data octet in this segment
except when the SYN flag is set. If SYN is set, this field contains the initial sequence
number (ISN) and the first data octet in this segment has sequence number ISN + 1.

28
Acknowledgment Number (32 bits): Contains the sequence number of the
next data octet that the TCP entity expects to receive.
 Data Offset (4 bits): Number of 32-bit words in the header.
 Reserved (4 bits): Reserved for future use.
 Flags (6 bits): For each flag, if set to 1, the meaning is
CWR: congestion window reduced.
ECE: ECN-Echo; the CWR and ECE bits, defined in RFC 3168, are used for the
explicit congestion notification function; a discussion of this function is beyond our
scope.
URG: urgent pointer field significant.
ACK: acknowledgment field significant.
PSH: push function.
RST: reset the connection.
SYN: synchronize the sequence numbers.
FIN: no more data from sender.

29
The Sequence Number and Acknowledgment Number are
bound to octets rather than to entire segments. For example, if a
segment contains sequence number 1001 and includes 600 octets
of data, the sequence number refers to the first octet in the data
field; the next segment in logical order will have sequence
number 1601.

30
UDP transmits segments consisting of an 8-byte header followed by the
payload. The header is shown in Fig. 6-23. The two ports serve to
identify the end points within the source and destination machines. When
a UDP packet arrives, its payload is handed to the process attached to the
destination port.

31
 Congestion Control
Regulating the Sending Rate
When the load offered (offered traffic) to any network is more than it can handle,
congestion builds up.

When a connection is established, a suitable window size (amount of data can be sent
without acknowledgement, initially size of the TCP packet) has to be chosen. The receiver
can specify a window based on its buffer size.

If the sender sticks to the receiver’s window size, problems will not occur due to buffer
overflow at the receiving end, but they may still occur due to internal congestion within
the network.

32
 In Fig.(a)-(b), we see this problem illustrated hydraulically. 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 network.
If too much water comes in too fast, it will back up and some will be lost (in this case by overflowing the
funnel).

Figure (a) A fast network feeding a low-capacity receiver. (b) A slow network feeding a high-capacity receiver.
33
The Internet solution is, to realize that two potential problems : network
capacity and receiver capacity and to deal with each of them separately.
To do so, each sender maintains two windows: the window the receiver
has granted and the congestion window.

 The number of segments to transmit, TCP uses a variable called a congestion window,
cwnd, whose size is controlled by the congestion situation in the network.

 Another variable or window rwnd related to the congestion at the end (related to buffer
flow like sliding window of LLC).

 The actual size of the window is the minimum of these two.

Actual window size = minimum (rwnd, cwnd)

34
Because the spare room changes with time, rwnd is dynamic.

The receive window (rwnd)and the receive buffer (RcvBuffer)

Relevant to rwnd Relevant to cwnd

35
TCP Congestion Control

36
 Fig. below shows what happens when a sender on a fast network (the 1-Gbps link)
sends a small burst of four packets to a receiver on a slow network (the 1- Mbps link)
i.e. the bottleneck or slowest part of the path.

 Initially the four packets travel over the link as quickly as they can be sent by the sender.
At the router, they are queued while being sent because it takes longer to send a packet
over the slow link than to receive the packet over the fast link.

Low data rate elongates the packets

37
 Eventually the packets get to the receiver, where they are acknowledged. The times for
the acknowledgements reflect the times at which the packets arrived at the receiver after
crossing the slow link.

 These acknowledgements travel over the network and back to the sender they preserve
this timing.

A burst of packets from a sender and the returning ack clock

38
 The key observation is this: the acknowledgements return to the sender at about the
rate that packets can be sent over the slowest link in the path. This is precisely the
rate that the sender wants to use.
 If it injects new packets into the network at this rate, they will be sent as fast as the
slow link permits, but they will not queue up and congest any router along the path.
This timing is known as an ack clock. It is an essential part of TCP.

 By using an ack clock, TCP smoothens out traffic and avoids unnecessary queues at
routers.

A burst of packets from a sender and the returning ack clock

39
Congestion Window
 TCP maintains a new state variable for each connection, called Congestion
Window, which is used by the source to limit how many TCP packets (or TCP
segments) is allowed to transmit at a given time.

 The unit of Congestion Window is number of packets can be sent without ACK
like sliding window of data link layer.

Here we consider two congestion control algorithms graphically:

TCP Tahoe and TCP Reno.

40
 Internet congestion control algorithm of TCP Tahoe
As illustrated in Fig. below, the maximum size
of segment is: 1 segment = 1024 bytes = 1KB.

During slow start, at round r = 0, the number of

TCP segment sent and wait for ACK is 1 = 1KB.
Slow start
If all the segments arrived (after RTT), then at
round r = 1, the number of TCP segments sent
and wait for ACK is 2 = 2 KB.

Similarly, at round r = 2, the number of TCP

segment sent and wait for ACK is 4 = 4KB.

The slow start will increase the TCP segments The slow start, but it is not slow at all;
as: 1, 2, 4, 8, 16 till the threshold value 32 it is exponential growth.
(decided before).
41
 After round r = 5, the size of window Slow start/ exponential
reaches at threshold, then on the increase.
subsequent round the algorithm will enter
additive increase (linear) mode i.e. 33,
34, 35, … till experiences loss of packet. Additive increase or linear
increase or congestion
 At round, r = 13, the network experience avoidance phase.
first loss of TCP packet, with size of
window of 40. Therefore it will go to the
initial state i.e. at r = 14, the size of
window will be 1 = 1KB and begins slow
start. Multiplicative
decrease.
 Now the threshold value of exponential
increasing part will be just half of
previous case i.e. 40/2 = 20 KB.

 At round r = 19, it will enter linearly

increasing state and will continue till
experiencing loss of packet.
42
TCP Tahoe Parameters

43
The threshold value of congestion window is the maximum allowed value of the
congestion window till which the window grows exponentially.

Initially, the threshold congestion window was 64 KB, but a timeout occurred, so the
threshold is set to 32 KB and the congestion window to 1 KB for transmission round r = 0
here. The congestion window then grows exponentially until it hits the threshold (32 KB).
After reaching threshold, it grows linearly like 33, 34, 35, ….

Slow start followed by additive increase in TCP Tahoe 44

If no more timeouts occur, the congestion window will continue to grow up to the
size of the receiver's window. At that point, it will stop growing and remain constant
as long as there are no more timeouts and the receiver's window does not change size.

If the loss is detected by the triple dup ACKs, the lost packet is retransmitted, the
threshold is set to half the current window and slow start is initiated all over again.

Triple duplicate ACK

Packet n is lost, but packets n+1, n+2, etc. arrive
Receiver sends duplicate acknowledgments and the sender
retransmits packet n quickly
Do a multiplicative decrease and keep going
45
 RTT→ Round Trip Time

Slow start from an initial congestion window of one segment

46
 RTT→ Round Trip Time

Additive increase from an initial congestion window of one segment

47
In TCP Reno, instead of repeated slow starts, the congestion window of a running
connection follows a sawtooth pattern of additive increase (by one segment every RTT)
and multiplicative decrease (by half in one RTT).
RTT→ Round Trip Time

Fast recovery and the sawtooth pattern of TCP Reno

48
Example-1
Figure on next slide shows an example of congestion control in a Taho TCP version. TCP starts with slow-start
(SS) threshold, ssthresh of 16 MSS (maximum segment size ). TCP begins at the slow-start (SS) state with cwnd
= 1. The congestion window grows exponentially, but a time-out occurs after the third RTT (before reaching the
threshold). TCP assumes that there is congestion in the network. It immediately sets the new ssthresh = 4 MSS
(half of the current cwnd, which is 8) and begins a new slow-start (SA) state with cwnd = 1 MSS. The
congestion grows exponentially until it reaches the newly set threshold. TCP now moves to the congestion-
avoidance (CA) state, and the congestion window grows additively until it reaches cwnd = 12 MSS. At this
moment, three duplicate ACKs arrive, another indication of the congestion in the network. TCP again halves the
value of ssthresh to 6 MSS and begins a new slow-start (SS) state. The exponential growth of cwnd continues
till the window reaches the ssthresh (6) and the TCP moves to the congestion-avoidance state. The data transfer
now continues in the congestion-avoidance (CA) state until the connection is terminated after RTT 20.

49
50
This type of TCP congestion control is often referred to as an additive-increase,
multiplicative decrease (AIMD) form of congestion control. AIMD congestion control gives
rise to the “saw tooth” behavior shown in Figure below.

Additive-increase, multiplicative-decrease congestion control

51