In the last decade, cellular networks have experienced an enormous popularization, being even much higher than the initial most optimistic forecasts. The interest aroused among users has made cellular networks to become the most profitable sector in telecommunications. Although the number of users is not expected to increase due to the fact that the number of mobile lines is higher than the population in many countries, there exists a wide range of new challenges for operators in order to offer attractive and competitive services to users. These new services are expected to demand also a higher amount of resources. For the research community, this calls for the necessity of developing more efficient and complex resource management mechanisms in order to guarantee some requirements of quality of service. Traditionally, to design these resource management mechanisms the starting point has been their corresponding proposals for fixed networks. Notwithstanding, cellular networks add some new challenges due to the scarcity of the spectrum, the randomness in propagation and the mobility of terminals.

The framework of this thesis are cellular networks that implement admission control policies. More concretely, in the first part of the thesis some analytical cellular models have been proposed to study the impact of retrials produced in a system when the admission controller blocks an access request to the system. This contribution is done in both the development of useful techniques for solving retrial systems and in the application of those techniques to cellular network models to understand the impact that retrials have on performance. The second part of the thesis is devoted to the design of optimal admission control policies. Since in current cellular networks it is possible to estimate the future behaviour of terminals from information like its position, velocity, historic behaviour, location in a road map... we have incorporated this information into the admission control optimization process.