In the last decade data traffic of mobile networks has grown significantly due to the introduction of new generation mobile networks. If the increasing data traffic is not handled properly, the utilization and quality of the network deteriorates, as well as the satisfaction of the users. Therefore it is important to work out methods that help the design process and the operation of the network.
During the evolution of mobile networks the size of the cells were constantly decreasing but at the same time the number of the so called handoff or handover messages, which arise during stepping across cell borders, gradually increased. One has to be prepared to receive incoming demands in order to reserve resources for the message. Based on previous movements of users their future location can be predicted. The effectiveness of mobility prediction methods therefore determines the QoS parameters of the network.
The aim of my thesis was to develop an algorithm that was able to predict effectively the number of users in a given network, preparing service providers for the incoming demand. This model can be used primarily in network situations where the cells are crossed by a road with heavy traffic that siginficantly modifies the behaviour of users.
In the first part of the thesis I will demonstrate and contrast mobility models found in the literature and later in chapter 4.4 and 4.5 I will introduce the ring based call admission control algorithms that I developed. Later I will compare and contrast them and will evaluate the results.