Genetic search for optimal personalized location areas for future PCS networks

In Personal Communication Services (PCS) networks, mobile terminals (MT) are communicating frequently using a variety of services. A high-quality framework is needed for tracking consistently MTs in the system and delivering the services timely. Future PCS networks are expected to have smaller cells to address the transmission power concerns and reuse frequencies efficiently. Thus, MTs are expected to signal even more frequently under always update (AU) strategy adopted in current PCS networks, such as GSM . One pathway for solving the problem is the consideration of Location Areas (LA). The location update takes place only when the MT moves off the area. For call delivery the area is paged. Thus, another signaling component is introduced. In this paper, we propose the design of personalized LA's (PLAs) per an individual user. The model we use takes into account the MT's behavior in the network and the system parameters. The problem is known to be NP-complete of a high computational complexity. We provide a proof of the complexity and solve the problem by using the genetic search for finding a sub-optimal solution. It is shown that the cost of location management under the model and the designed search algorithm can be significantly lower than that consumed by AU mechanism.