Circuit-switched and packet-switched networks are types of telecommunication networks. Techspirited explains and compares these two types of networks.
Apt Examples
The old telephone system (landline telephone system) is an example of circuit-switched networks. The Internet is an example of packet-switched network based on the TCP/IP protocol.
Telecommunication is the practice of transmitting messages electronically between two points separated over a distance. These messages are sent across a telecommunication network. It consists of terminal nodes, intermediate nodes, and links that connect the nodes. Circuit- and packet-switched networks are used to send a message across between two points.
Circuit-switched Networks
Here, communication between two nodes happens by establishing a dedicated communication route between them. This route of communication is established through a sequence of links within the network. These links start from the source and travel from one node to another and end at the destination. So, how does communication by circuit switching happen?
✦ Step 1: To communicate a message from one point to another, the first thing that we need to do is to connect them. This is done by logically establishing a circuit between the two points. In this circuit, some components are dedicated while some are shared.
✦ Step 2: Now we have to actually send the message across from one point to another. Mostly, a full-duplex connection is established for this. In the circuit-switching technique, we send either an analog or digital signal depending on the nature of the network.
✦ Step 3: Once the message has been sent, we have to terminate the connection. This is done by sending signals to free the resources.
Within the mesh of the network, we figure out a way to connect the source and destination, and then physically connect them through interlinking nodes.
Packet-switched Networks
Here, data to be transferred is broken down into small subsets known as packets. But there is no fixed path (or circuit) between the source and the destination through which the entire message is communicated. Instead, every packet of the message follows a different route to reach the final destination. Each packet has a header that contains the necessary routing information. Only one dedicated connection between the source and the destination is not set up. This approach has the advantage that during the instant when there is no communication happening between one pair of sender and receiver, the line can be used by another pair. At the receiving end, these packets are then rearranged in their proper sequence (this information is contained in the headers) and the entire message is constructed. There are two basic ways of packet switching:
Virtual Circuit Packet Switching In this technique, initially there is a phase known as the setup phase, in which a path between the source and the destination is established. All the intermediate nodes that connect the source and the destination are determined before the packets of the message are sent across.
During the initial setup phase, every intermediate node is given the information about the route of the packets in the form of a table. The packets that pass through these intermediate nodes have short headers that store the information about the next node, and not the final destination. Common examples of Virtual Circuit Networks include X.25 and Frame Relay that are used in public data networks (PDN).
Datagram Packet-Switching In this technique, every packet is considered as a separate, independent unit. The header of every packet has the complete information about where it has to finally reach. At every in-between node, the decision about where a packet has to go next is taken. This decision is taken considering:
- The shortest possible path that can be taken to reach the final destination.
- The availability of the next free node through which the packet can travel.
Protocols such as Routing Information Protocol (RIP) and Open Shortest Path First (OSPF) are put to use while determining the shortest possible path. Hence, it is a dynamic, decide-on-the-go technique of passing packets. However, in this technique, since every packet follows a different route to reach the ultimate address, it may happen that they arrive there in a different order. The packets then have to be sorted to put together the entire message.
The Real Comparison
Circuit-switched Networks | Packet-switched Networks |
The entire message is passed. | The message is broken down into small parts called packets. |
The message is passed by establishing a dedicated communication link between them. | There is no dedicated communication line. The individual packets are passed independently through intermediate nodes by the shortest route available. |
Only one message passes through any instant. | The communication line is shared by several packets by waiting in a queue. |
There is lesser overhead while communicating the message. | Each packet has a header that contains the necessary routing information. Hence, overheads are associated during transmission. |
The transmission of message happens comparatively faster. | Sometimes delays can occur while the message is being transmitted. At the receiving end, these packets are then rearranged in their proper sequence (this information is contained in the headers), and the entire message is constructed. |
Traditionally, voice data has been sent over a packet-switched network. Circuit-switched networks are used in Public Switched Telephone Network (PSTN), which is the old kind of telephone system to transmit voice signals. | Earlier voice data was not sent over a packet-switched network. But now it is used in VoIP (Voice Over Internet Protocol). [VoIP is a service in which you can use the Internet to make phone calls.] |
Sending binary data is not feasible over a circuit-switched network. | It is more feasible to send binary data over a packet-switched network. |
If a particular part of the message is damaged, the whole file has to be resent. | If a particular part of the message is damaged, only the damaged packet has to be resent. |
If there is a “dead time” or “silent time” when no communication is happening, the resources are locked and the communication line cannot be used by another sender. | When no communication happens in this case, the resources are not locked up and the communication line can be used by another sender. |
If a dedicated circuit is not required, a circuit-switched network may not be economical. | A packet-switched network may be more economical when you don’t require a dedicated network. |
It is more reliable than packet-switched networks as there is no possibility of packets being lost or arriving out of order. | It is not reliable as there is a possibility of packets being lost or arriving out of order. |