Unit 8: Network Layer 5.

Error Control: Error Control is performed end to end in this layer to ensure that
8.1 Transport layer issues; Congestion control; Flow control; Quality of the complete message arrives at the receiving transport layer without any error.
Error Correction is done through retransmission.
services:
Transport layer issues:
 The transport layer is a 4th layer from the top.
 The main role of the transport layer is to provide the communication services directly
to the application processes running on different hosts.
 The transport layer provides a logical communication between application processes
running on different hosts.
 The transport layer protocols are implemented in the end systems but not in the
network routers.
 A computer network provides more than one protocol to the network applications. For
example, TCP and UDP are two transport layer protocols that provide a different set of Design Issues with Transport Layer
services to the network layer.  Accepting data from Session layer, split it into segments and send to the network layer.
 Ensure correct delivery of data with efficiency.
 All transport layer protocols provide multiplexing/demultiplexing service. It also
 Isolate upper layers from the technological changes.
provides other services such as reliable data transfer, bandwidth guarantees, and delay
 Error control and flow control.
guarantees.
Congestion control:
Congestion is a situation in which too many sources over a network attempt to send data
and the router buffers start overflowing due to which loss of packets occur. As a result,
retransmission of packets from the sources increases the congestion further. In this
situation, the Transport layer provides Congestion Control in different ways. It uses open
loop congestion control to prevent the congestion and closed loop congestion control to
remove the congestion in a network once it occurred. TCP provides AIMD- additive
increase multiplicative decrease, leaky bucket technique for congestion control.
Flow control:
The transport layer provides a flow control mechanism between the adjacent layers of the
TCP/IP model. TCP also prevents data loss due to a fast sender and slow receiver by
Functions of Transport Layer imposing some flow control techniques. It uses the method of sliding window protocol
1. Service Point Addressing: Transport Layer header includes service point address which is accomplished by the receiver by sending a window back to the sender informing
which is port address. This layer gets the message to the correct process on the the size of data it can receive.
computer unlike Network Layer, which gets each packet to the correct computer. Quality of services (QoS)
2. Segmentation and Reassembling: A message is divided into segments; each QoS is an overall performance measure of the computer network.
segment contains sequence number, which enables this layer in reassembling the Important flow characteristics of the QoS are given below:
message. Message is reassembled correctly upon arrival at the destination and 1. Reliability
replaces packets which were lost in transmission. If a packet gets lost or acknowledgement is not received (at sender), the re-transmission of
3. Connection Control: It includes 2 types: data will be needed. This decreases the reliability.
 Connectionless Transport Layer: Each segment is considered as an independent The importance of the reliability can differ according to the application.
packet and delivered to the transport layer at the destination machine. For example:
 Connection Oriented Transport Layer: Before delivering packets, connection is E- mail and file transfer need to have a reliable transmission as compared to that of an
made with transport layer at the destination machine. audio conferencing.
4. Flow Control: In this layer, flow control is performed end to end. 2. Delay
Delay of a message from source to destination is a very important characteristic. However,
delay can be tolerated differently by the different applications.
For example: software (protocol stack), such as an entry in a table (listing communication protocol,
The time delay cannot be tolerated in audio conferencing (needs a minimum time delay), destination, status, etc.), and is a form of system resource.
while the time delay in the e-mail or file transfer has less importance. The term socket is analogous to physical female connectors, communication between two
3. Jitter nodes through a channel being visualized as a cable with two male connectors plugging
The jitter is the variation in the packet delay. into sockets at each node. Similarly, the term port (another term for a female connector) is
If the difference between delays is large, then it is called as high jitter. On the contrary, if used for external endpoints at a node, and the term socket is also used for an internal
the difference between delays is small, it is known as low jitter. endpoint of local inter-process communication (IPC) (not over a network). However, the
Example: analogy is limited, as network communication need not be one-to-one or have a
Case1: If 3 packets are sent at times 0, 1, 2 and received at 10, 11, 12. Here, the delay is dedicated communication channel.
same for all packets and it is acceptable for the telephonic conversation. 8.3 Segmentation and reassembly:
Case2: If 3 packets 0, 1, 2 are sent and received at 31, 34, 39, so the delay is different for Segmentation:
all packets. In this case, the time delay is not acceptable for the telephonic conversation. Segmentation is the term used to describe the process of chopping streams of data into
4. Bandwidth smaller chunks. Segmentation usually occurs fairly early in the communication process and
Different applications need the different bandwidth.
it is almost always software that performs the segmentation process. The segmentation
For example:
Video conferencing needs more bandwidth in comparison to that of sending an e-mail. process is performed prior to transfer of data across a network or before storage on a
8.2 Transport layer addressing, sockets: peripheral device. Segmentation is necessary because today's communication systems use
what is called packetized communication.
Transport layer addressing:
An example of segmentation would be when Transmission Control
In the same way that a single physical medium must be multiplexed at times, the protocol
Protocol (TCP of TCP/IP fame) chops an e-mail into a segments, encapsulates the segment
stack in a host must be multiplexed, or shared, by many application processes at once. This
means the network layer address (e.g. IP address) isn't sufficient to identify the process with remote and local TCP port numbers and then delivers the completed protocol data unit
using a transport layer connection (e.g. a socket). Tanenbaum calls these end point address is passed from TCP to Internet Protocol (IP) to be stamped with a sequence number, source
TSAP (transport service access points). One way of seeing why a unique TSAP is and destination addresses added and a checksum calculated.
necessary is to ask yourself the question, "to which process does a transport TPDU get Reassembly:
delivered when it arrives?" Reassembly is the reverse of segmentation. Protocol Data Units are put back together in the
In the Internet suite the multiplexing of TCP segments is done by port numbers. Each correct order to reassemble a stream of data in its original form.
process using a socket is given a unique (to that machine) port number. Then the 8.4 Connection oriented and connectionless service:
combination of IP address + port number is a unique identifier for that process. Connection Oriented Services:
How does a client process know the TSAP of a server process? Some possibilities There is a sequence of operation to be followed by the users of connection-oriented service.
These are:
1. TSAPs are well-known, documented, static. Works ok for basic sorts of services 1. Connection is established.
that all servers provide. Example would be Internet well-known services (ports 2. Information is sent.
below 1024). 3. Connection is released.
2. A TSAP server runs on a well-known TSAP. A protocol is used for clients to query In connection-oriented service we have to establish a connection before starting the
the TSAP server to find the TSAP given a name for the service. This is like a name communication. When connection is established, we send the message or the information
server. The services are identified by string names, the name server returns the and then we release the connection.
TSAP (e.g. a port number). Connection oriented service is more reliable than connectionless service. We can send the
3. A meta-server is used for the initial connection. The meta-server (server, or super message in connection-oriented service if there is an error at the receiver's end. Example of
server) answers requests on many TSAPs, starts the proper server process, then connection oriented is TCP (Transmission Control Protocol) protocol.
gives the incoming connection to the new process. This also saves having to launch Connection Less Services
many infrequently used server processes. The Internet INETD (Internet Service It is similar to the postal services, as it carries the full address where the message (letter) is
Daemon) is an example of this. to be carried. Each message is routed independently from source to destination. The order
Sockets: of message sent can be different from the order received.
A network socket is an internal endpoint for sending or receiving data within a node on In connectionless the data is transferred in one direction from source to destination without
a computer network. Concretely, it is a representation of this endpoint in networking checking that destination is still there or not or if it prepared to accept the message.
Authentication is not needed in this. Example of Connectionless service is UDP (User 3. Flow Control:
Datagram Protocol) protocol. If sender sends data faster than what receiver can process then the receiver will drop
Difference: Connection oriented and Connectionless service the data and then request for a retransmission, leading to wastage of time and
1. In connection-oriented service authentication is needed, while connectionless resources. TCP provides end-to-end flow control which is realized using a sliding
service does not need any authentication. window. The sliding window sends an acknowledgement from receiver’s end
2. Connection oriented protocol makes a connection and checks whether message is regarding the data that the receiver can receive at a time.
received or not and sends again if an error occurs, while connectionless service UDP doesn’t implement flow control and depends on the higher layer protocols for
protocol does not guarantee a message delivery. the same.
3. Connection oriented service is more reliable than connectionless service. 4. Connection Oriented:
4. Connection oriented service interface is stream based and connectionless is TCP is connection oriented, i.e., it creates a connection for the transmission to take
message based. place, and once the transfer is over that connection is terminated.
8.5 Transport layer protocols: TCP, UDP UDP on the other hand is connectionless just like IP (Internet Protocol).
Layer 3 or the Network layer uses IP or Internet Protocol which being a connection less 5. Reliability:
protocol treats every packet individually and separately leading to lack of reliability during TCP sends an acknowledgement when it receives a packet. It requests a
a transmission. For example, when data is sent from one host to another, each packet may retransmission in case a packet is lost.
take a different path even if it belongs to the same session. This means the packets UDP relies on the higher layer protocols for the same.
may/may not arrive in the right order. Therefore, IP relies on the higher layer protocols to
provide reliability.
TCP (Transmission Control Protocol):
TCP is a layer 4 protocol which provides acknowledgement of the received packets and is
also reliable as it resends the lost packets. It is better than UDP but due to these features it
has an additional overhead. It is used by application protocols like HTTP and FTP.
UDP (User Datagram Protocol):
UDP is also a layer 4 protocol but unlike TCP it doesn’t provide acknowledgement of the
sent packets. Therefore, it isn’t reliable and depends on the higher layer protocols for the
same. But on the other hand, it is simple, scalable and comes with lesser overhead as
compared to TCP. It is used in video and voice streaming.
TCP Vs UDP –
1. Session Multiplexing:
A single host with a single IP address is able to communicate with multiple servers.
While using TCP, first a connection must be established between the server and the
receiver and the connection is closed when the transfer is completed. TCP also
maintains reliability while the transfer is taking place.
UDP on the other hand sends no acknowledgement of receiving the packets.
Therefore, provides no reliability.
2. Segmentation:
Information sent is first broken into smaller chunks for transmission.
Maximum Transmission Unit or MTU of a Fast Ethernet is 1500 bytes whereas the
theoretical value of TCP is 65495 bytes. Therefore, data has to be broken into smaller
chunks before being sent to the lower layers. MSS or Maximum Segment Size should
be set small enough to avoid fragmentation. TCP supports MSS and Path MTU
discovery with which the sender and the receiver can automatically determine the
maximum transmission capability. UDP doesn’t support this; therefore it depends on
the higher layer protocols for data segmentation.