Transport Layer

Layer 4 of the OSI reference model is the Transport layer. The transport layer is primarily responsible for the message’s process-to-process delivery. A process is a type of application program that runs on the host computer.

The Transport layer’s primary role is to collect data from the layer above, divide it into smaller units, transfer these data units to the Network layer, and verify that all of the parts arrive at the other end in working order.

Furthermore, all of this must be accomplished quickly and in a way that protects the upper layers from changes in hardware technology.

The Transport layer also decides what kind of service to offer the Session layer and, eventually, the network’s users. An error-free point-to-point channel that transmits messages or bytes in the order they were sent is the most common sort of transport link.

From the source to the destination, the Transport layer is a genuine end-to-end layer. In other words, by utilizing message headers and control messages, a program on the source machine communicates with a program on the destination machine.

Transport Layer Elements

This layer’s Elements comprise the following.

Service-point Addressing

Because computers usually execute multiple programs at the same time, delivering data from source to destination using two computers is not feasible. As a result, this layer is used to add a specific type of address to its header, referred to as a port address or service point address. Every packet is delivered to the correct computer via this address, and the transport layer delivers the entire message to the correct method on that machine.

Segmentation & Reassembly

A message can be segmented into communicable fragments, with each segment containing a sequence number that allows the layer to reconstruct the message. Once the message has arrived at the target system, it can be reconstructed properly, with packets recognized and replaced as needed due to lost data during transmission.

Connection Control

Connectionless and connection-based are the two categories. The transport layer in connectionless treats each packet as an individual and sends it to the target computer. The receiver does not transmit an acceptance to the sender regarding a packet receipt in this type of communication. This is an older mode of communication.

Connection based Transport Layer

Before transmitting packets to the target, this layer establishes a connection with the layer at the computer’s destination. The following steps can be used to establish the connection:

  • Establishment of Connection
  • Data Transfer
  • Connection Termination

As a result, connection-based service is used for all data transmissions, but connectionless service is less reliable.

Multiplexing

Several packets are sent across a system by control systems from various applications, and these are found within the transport layer. This layer permits packets to be sent in a variety of ways, and these packets are identified by their port numbers, which are then transmitted to the layer after the relevant headers have been added.

De-multiplexing

The data that is transferred from distinct processes can be acquired through the de-multiplexing procedure. It obtains data segments from this layer in order to send them to the appropriate method operating on the receiving system.

Flow Control

The transport layer is also in charge of flow regulation between the TCP/IP model’s contiguous layers. Even if it performs a continuous node, it does not execute across a single connection. Data loss from the sender and slow receiver can be prevented using excellent flow control approaches.

Error Control

This is likewise accomplished in a back-to-back manner, much like the data link layer. The entire message can appear at the receiving end of the layer without any errors in this layer. The packet re-transmission can be used to improve the error rate. Once the data has arrived, the integrity of the data can be confirmed by using services like the ACK and NACK to inform the sender.