Why are mobile networks 'cellular'?

Everybody is familiar with the word 'cellular'. But, how many of you can actually explain what a cellular system is? And most importantly, can you explain why all mobile technologies are structured this way? In this first post I will try to enlighten you about this topic as a basis for future posts in which I will deal with topics that require a deep understanding of mobile communications.

Mobile technologies, no matter the generation, use the air interface as their transmission medium. More specifically, it is the electromagnetic spectrum what is used to transmit modulated waves containing information. Since electromagnetic spectrum is a very limited resource, and only a little portion of it is reserved for mobile communications, it is necessary to implement some sort of multiplexing in order to maximize the number of simultaneous users. To achieve this, the transmission bands are divided into channels.

GSM bands.

For example, in GSM-900, transmissions can only take place between 890-915 MHz for the uplink, and 935-960 MHz for the downlink. Each of these bands is divided into channels of 200 KHz width. This would only allow 124 concurrent users in the network, which obviously is not enough. This number can be increased by applying time multiplexing (just dividing each frequency channel into time slots), but the result is still far from optimal.

Additionally, it has been said that the channels are 200 KHz width, and this is not strictly true. The fact is that 99% percent of the transmitted power is contained inside a 250 KHz width band. Since the actual channel spacing is 200 KHz, two adjacent GSM channels interfere with each other, making impossible to use these two channels simultaneously. It is pretty obvious this system is not efficient at all.

Public mobile networks must be able to provide service to a large number of users. To cope with the before mentioned problems, a cellular scheme is adopted. This architecture involves dividing the total coverage area of the network into smaller regions or 'cells', each of which is assigned a set of non-adjacent transmission channels different from the ones assigned to neighboring regions. Each cell is controlled by a network entity (i.e BTS for 2G, NodeB for 3G) which is responsible for managing the cell resources.

The main advantage of using a cellular system is the frequency reuse. For a given coverage area, two or more sufficiently separated cells can operate in the same channels without interference. This sort of 'spatial multiplexing' drastically improves the performance of the network, allowing multiple users to share the same resources at the same time. Normally, cells are grouped into clusters, which are frequency distribution patterns that are replicated all over the total coverage area of the network.

7-cell clusters. f_i denotes a set of frequencies.

The minimum reuse distance depends on the cell radius and the number of cells per cluster, which typically is 4, 7, 12 or 21. This is a key factor, because the smaller the number of cells per cluster is, the bigger the cell capacity becomes, since there are more channels per cell. On the other hand, smaller clusters are more susceptible to suffer interference from neighboring clusters, so the size has to be balanced.

Cells can be classified regarding to the extension of their coverage area. Macrocells, for example, provide service to large rural areas, while microcells, picocells and femtocells are common in densely populated urban environments.

As a first approach to cellular telephony, this is all for now. In the next post I will focus on the cell concept, explaining its elements, and the logical network divisions derived from it.