What is a Packet Switching : Modes and Delays

We live in an era where every process is quick and responsive. Packet switching is one such technology used nowadays on data networks like the internet, LAN, WAN. The development of Voice over Internet Protocol (VoIP), enabled packet switching to easily transmit data and voice traffic. This allowed businesses to experience huge benefits in terms of cost, efficiency, and scalability. This switching refers to routing and transmitting of data in an efficient manner through the channel/network with minimal latency. The channel is made available for other purposes upon the completion of the transmission.

What is Packet Switching?

Definition: Packet switching refers to a set of protocols that uses a connection-less network switching approach to transmit the packets. In this switching, the messages are broken and grouped into small units called packets. These packets are transmitted individually across a digital network to reach its destination. Packets need not follow the same route to reach their destination. As all the packets arrive at a destination in a different order, the original message is recompiled by the destination itself. The packet switching diagram is shown below.

In this switching, packets have two parts – a header and a payload. The information in the header allows networking hardware/intermediate node to make sure that the packets are directed towards its destination, while the definite data is carried by the payload.

Each packet has a source and destination address to travel independently over a network with a variable bitrate. Packets are forwarded asynchronously by intermediate nodes because of congestion, queuing, and so on, and hence follows different routes. These packets arrive at the destination in a different order, and the destination ensures to reassemble the data of the same file.

The message consists of four packets – A, B, C, and D. Each packet consists of source and destination address and follows more than one route to reach the destination from source as shown in the figure below.

Modes of Packet Switching

Packet switching is classified into two main types. They are:

Connection-oriented Packet Switching

This is also known as virtual circuit switching, which requires a setup phase or virtual connection to establish a path before transmission. With signaling protocol, a predefined path is set up to enable the sender, receiver and all packets of the same message to follow this path. Switches/routers provide virtual circuit ID to recognize the virtual connection. The data in this type of switching is split into small units. A sequence number is added to these small units. In this process, three phases are described. They are set up, data transfer, and tear down phase.

In the setup phase, address information is only transferred to each node. As soon as the route to the destination is found, an entry is added to the switching table of each intermediate node.

In the data transfer phase, the packet header may contain information like length, timestamp, and sequence number. This information can be different for different packets.

One of the prominent applications of connection-oriented packet switching is in the switched WAN. Protocols like X.25, Frame-Relay, ATM (asynchronous transfer mode), and multi-protocol label switching use this type of switching approach.

Connectionless Packet Switching

Connectionless type Switching is popularly known as datagram switching. Here, each packet comprises a source and destination address and port address and other necessary information. Sometimes, the packets are labeled with a sequence number.

In datagram packet switching, the packets traverse independently and in different routes and so the packets that arrive at the destination might be out-of-order delivery. As packets arrive at the destination in an unordered format, the original message shall be retrieved based on the sequence numbers of the packets.

Reliable delivery of packets in connectionless switching is not guaranteed. So, providing end-to-end systems with additional protocols is needed.

Delays in Packet Switching

The four types of delays in this switching are:

Transmission Delay

It simply refers to the time taken to send out all the packets, or, time is taken to absorb all the data bits onto the communication medium. Transmission delay relies on the length of the packet and the network’s bandwidth.

Transmission delay = data size/ bandwidth = (L/B) second

Propagation Delay

Propagation delay refers to the time taken by the bits to travel from source to destination over the link. Distance and propagation speed are the factors that impact propagation delay.

Propagation delay = distance/transmission speed = d/s

Queuing Delay

Queuing delay occurs because of the nature of traffic in the network. Therefore, it refers to the time spent waiting in a queue until it gets executed and is defined as follows –

Average queuing delay = (N-1) L/ (2*R)

Where ‘N’ is the no. of packets

‘L’ is the size of the packet

‘R’ is the bandwidth

Processing Delay

It refers to the time taken to process a packet. Processing delay also refers to the time required to check for bit errors, determine output link, etc.

Total time or End-to-end time = Transmission delay + Propagation delay + Queuing delay + Processing delay

Advantages of Packet Switching Over Circuit Switching

This switching offers various benefits compared to circuit switching and they are listed below:

• It delivers the data to a destination by finding their own paths; circuit switching has a dedicated and predefined channel.
• It is highly reliable as missing packets are detected by destination; circuit switching does not have this option.
• It uses lesser bandwidth as packets are quickly routed towards the destination; circuit switching should have dedicated bandwidth.
• The channel in this switching is available for other transmissions as soon as packets are routed; circuit switching occupies the channel till the voice communication is completed
• It is cost-effective and easier to implement; circuit switching is expensive

Disadvantages of Packet Switching Over Circuit Switching

Despite offering various benefits, this switching offers disadvantages too, which are listed below:

• As the movement of packets is not synchronous in this switching, it may not be suitable in communication applications like voice calls; while circuit switching is highly suitable for voice calls
• Packets don’t move in an organized way, sequence numbers should be provided to identify each packet; circuit switching gives the highest priority for the channel to give the best experience to the users
• In this switching, complexity is high at each node as packets are routed over multiple paths to reach the destination, leading to loss of data or delay in delivering the packets; circuit switching makes sure there is no loss of data
• This switching needs additional and secure protocols to protect the data, leading to a significant rise in implementation costs; circuit switching has a dedicated channel for one service and one individual route.

FAQs

1). What is data packet switching?

The data packet switching is an approach used to transfer the data over a network in the form of packets. The data is broken into small units of variable length known as packets. Each packet having data in it travels along with the network..

2). Who invented packet switching?

The American scientist ‘Paul Baran’ explored the concept of packet switching in 1960. In 1965, Donald Davies developed a similar routing concept and named it as packet switching.

3). What are the switching techniques?

There are three types of switching techniques – Packet Switching, circuit Switching, and Message Switching.

4). What do you mean by switching?

Switching is a type of technique through which nodes can control or switch data to ensure that it is transmitted between given points on a network.

5). What is connectionless packet switching?

Connectionless Packet Switching is popularly known as datagram switching. Here, the message is broken and divided into packets. Each packet has a source and destination address to travel independently over a network. Packets are forwarded asynchronously by intermediate nodes because of congestion, queuing, and so on, and hence follows different routes. These packets arrive at the destination in a different order, and the destination ensures to reassemble the data of the same file.

Thus, in this article, we have discussed the concept of packet switching. The two techniques of packet switching with various advantages and disadvantages are discussed allowing the reader to comprehend which would be the best technology to be utilized for continuous and effective communication. A simple example of modern-day packet switching is email and Web pages using WAN and normal telephone service is considered as an example of circuit switching technology.