Make notes in this journal.

2.5 and 2.6 Circuit Switching vs packet switching

Using circuit switching a connection is not permanent and must be set up each time a call is made. The line is then used exclusively for that call and the network resources cannot be used for anything else for the duration of the call.

The limitations of circuit-switching which Davies tried to overcome were:

• It was slow compared with the operating speeds of a computer.
• It was expensive because the circuit was maintained for duration of call.

Although this was acceptable for voice calls it was wasteful for data communications which usually take place in short intense bursts with long gaps.

Packet switching, on the other hand made better use of the available space on the network. Data was broken down into small chunks that were labelled to show both the originating and destination addresses. The packets were forwarded from one computer to another until they arrived at their destination. If they were lost, they were automatically re-sent be the originator. The recipient acknowledged receipt of the packets as they arrived to eliminate unnecessary re-transmissions.

As the data being transmitted was in small packets it was able to be interleaved with other packets, from other sources, as it travelled through the networks. The route the packets took would be determined by each node as the packet reached it. This meant that the network capacity was better utilised and the system more efficient and cheaper to run.

Diagram from T171 site

back to essay