The personal communication service (PCS) is a system that allows communication anywhere in the world. PCS networks support mobile terminals (MTs) that are free to travel, and the network access point of an MT changes as it moves around the network coverage area. As a result, location management is important to effectively keep track of MTs with reduced signal flows and database queries. Even though dynamic location management strategies are known to show good performance, the static location management strategy which is easy to implement is considered in Chapter 2. A system with single home location register and pointer forwarding is assumed. An MT is assumed to have memory to store the IDs of visitor location registers (VLRs) each of which has the forwarding pointer to identify its current location. To obtain the registration point which minimizes the database access and signaling cost from the current time to the time of power-off, probabilistic dynamic programming formulation is presented. A Selective Pointer Forwarding scheme is proposed which is based on one-step dynamic programming. The proposed location update scheme determines the least cost temporary VLR which point forwards the latest location of the mobile. The computational results show that the proposed scheme outperforms IS-41, pure Pointer Forwarding, and One-step Pointer Forwarding at the expense of small storage and a few computations at the mobile terminals.
In Chapter 3, a new dynamic location management strategy is proposed. Several dynamic location management methods have been suggested to improve the efficiency of location management strategy. Among them, the distance-based scheme has been proved to have less mobility management cost than the others. However, the configuration of registration areas (RAs) for an MT needs to be matched with user’s movement behavior, street layout, and local topology. Thus, a probability-based location management strategy is proposed. In the proposed scheme, the time interval between two consecutive location registrations is divided into many small intervals and the RA is determined in each interval. The proposed scheme determines the shape of RA based on the movement behavior and the street layout. Furthermore, the size of RA in each interval is dynamically adjusted to minimize the expected cost. Transition probability matrix is used to implement the proposed strategy. The numerical results demonstrate that the proposed scheme significantly reduces mobility management cost compared to the distance-based strategy.
