Nowadays, various mobile services are gaining ground and they have both different burden on the network and different bandwidth requirements. The traffic of these services are served parallel, but it is important to ensure that necessary resources are provided for the priority services. It is a question in the third and fourth generation mobile networks, how to distribute the network link among services.
In the elastic traffic modeling, services are associated to traffic classes, where demands inside a class are receiving the same bandwidth and served in the same way. Weights are assigned to the classes and if the network link capacity is not enough to serve all demands in maximum speed, demands of lighter weight classes receive less bandwidth.
I created a transaction based simulation program to model elastic traffic. The simulator calculates characteristics of different traffic classes during running, including the expected value of serving time and the distribution of the number of the demands in the system. During the simulations I analyzed how these characteristics are affected by changing demand frequency, packet size, weight or maximum speed. I also analyzed how the simulation results differ from the theoretical values.