3.2 Segmentation of data

Data that is sent across a network is broken down into small chunks – or data units – first. This is important as switches and routers follow rules that require them to work with data units of a certain size and format.

The layered approach to the delivery of data across a network was introduced in Session 1. An analogy with shipping containers was used to explain this, where goods were packed into containers of the same size and then systems put in place to deal with containers of this size.

The four layers used to define the protocols (that is, the rules) that are used in networking are shown here.

At the top layer, data is created by users interacting with software. As this data is sent down the stack to the next layer – the Transport layer – it is broken down into units of the same size. This is called segmentation. Once the data has been segmented into data units, information is added to each one that is needed for this layer to be able to do its job. This information includes a unique identifier or address of the data unit’s destination.

By splitting data into smaller units, data from many different devices can share the network resources rather than them being dominated by a single data transmission. It also means if a data unit is corrupted, or lost, only that unit would be affected rather than the whole data stream.

Data units are called different things at different layers, as we will see shortly.

Once data units reach the Internet layer, this is where IP addresses come in. The Internet layer is responsible for getting data units across networks – sometimes via many different networks. The information that is added here includes the IP addresses of the source and destination networks. Routers work at this level and use IP addresses to get information from network to network. We will look at this in more detail in a later session.

You may have noticed that the information that was added at the layer above has now been included within the new data unit at the Internet layer. The Internet layer doesn’t need to know what information the layer above added: it is only concerned with its own information. This process of each layer adding its own information and creating its own data unit is called encapsulation.

At the Internet layer, data units become known as packets.

Finally, just before the data unit is launched onto the transmission medium – whether that be cables or a wireless link – it is passed to the Network Access layer. The Network Access layer is responsible for delivering data within the local area network, which could be another PC on the network itself or – if data is destined for outside the local network – it means getting it as a far as the networking device that will pass it onto other networks. As with the other layers, the data unit from the previous layer is encapsulated within a new data unit. The information that is added here includes the MAC addresses of the source and destination devices. Switches work at this level and use MAC addresses to get information across the local network – we will look at this in more detail next.

At the Network Access layer, data units become known as frames.