This paper discusses how a mobility model can be used jointly with a mobile activity trace and evolutionary computation to reduce the signaling load related to mobility management, an important and fundamental task in any public land mobile network. For this purpose, a mobility model is used to determine the most probable locations of each mobile subscriber, and this information, in turn, is used to assign paging areas. This paging strategy is evaluated by taking into account different probability thresholds and time-delay constraints, and in a multiobjective way. Thus, we study the whole objective space of the problem, ensure the results that are not dependent on the configuration of registration areas used in the analysis, and take into account the signaling traffic of both paging and location updates (in contrast to other published works, in which only the reduction in the paging load is considered). The feasibility of this paging scheme is evaluated by means of a performance analysis, in which it is compared with other paging schemes widely used in the recent literature. Results show that this paging strategy can reduce the blanket paging load by an average of ~56.73%. Furthermore, the performance analysis also shows that using evolutionary computation jointly with a paging procedure based on a mobility model is a very useful strategy for managing mobility in a public land mobile network, because it allows the total signaling load obtained by blanket paging to be reduced by ~67.03%.

WiMAX (Worldwide Interoperability for Microwave Access) or IEEE 802.16 is a technology developed to provide greater bandwidth and coverage. One of the most important challenges in wireless deploying network is the Base Stations (BS) location to meet traffic and coverage requirements efficiently. In this paper we are using a particularized swarm intelligence algorithm based on collective behavior of honey bees to find good solutions. In order to minimize the interference among cells and reduce energy consumption, we apply a post-processing to obtain power control and sectorization. Simulations results show that our method provides more efficient and better solutions than those are obtained by other evolutionary algorithms in the same instances.