Dynamic Planning of Personalized Location Areas for Future PCS Networks with a Simulated Annealing Algorithm

In current personal communication service (PCS) networks, such as Global Systems for Mobile Communications (GSM), the always-update (AU) strategy is used to keep track of mobile terminals (MTs) within the network. However, future PCS networks, like Universal Mobile Telecommunication System (UMTS) envision larger number of MTs and smaller cells. This causes frequent updating of locations, hence the increase of the signaling load. A proper location management scheme is crucial for the system effectiveness. In this paper, we present a dynamic location management scheme with personalized location areas (PLA). The proposed technology takes into account both the movement behavior of the individual users in the system and the system parameters. To analyze the location management cost, we model the network after a continuous time Markov chain. The PLAs are then dynamically defined for each MT such that the location management cost is minimized. The problem of dynamic planning of PLAs is proved to be NP-complete. Due to the high computational complexity of the problem, we search for an approximate solution using a simulated annealing (SA) algorithm. Simulation results show that the proposed dynamic location management scheme always outperforms the AU mechanism. Also the PLAs designed by the SA algorithm offer a lower signaling cost compared to that obtained using the heuristic algorithm