In this chapter our aim is to define a set of models used to evaluate the performances of systems and networks, based on the theoretical results and tools presented in the previous chapters. The essential objective is to provide as accurate a reflection as possible of the highly varied problems associated with modelling in an industrial environment (constructors, operators, laboratories, etc.).

The choice of these models is of course arbitrary and no attempt has been made to cover the whole of the subject. It is however based on two main criteria of essential importance:

• utility in our area of interest, that is the field of telecommunications;

• didactic interest from the viewpoint of modelling and the practical application of the theory. The level of demonstration will vary from one model to another, and it is the whole set of exercises which should provide a global view of modelling and resolution techniques. We will thus move from the simplest application of Markov graphs to the application of the Pollaczek method, while also covering simple rules of three and M/M/1 queues.

We will concentrate here solely on the field of traffic, as most of the important reliability models have already been covered in Chapter 6.

For the sake of clarity, we will draw a distinction between two major categories of models: those related to the control of systems and those to the field of information transport. Here also the distinction is arbitrary. It is primarily based on the most frequent usage in...