Probability criterion based location tracking approach for mobility management of personal communications systems

We consider a probability criterion based the location tracking scheme for personal communications systems. This scheme seeks to minimize the average signaling cost due to both paging and registration for individual mobile users. A timer-based location update and a single-step dynamic paging procedure are used. Each user-network interaction resets the timer and modifies the location record. The user registers when the timer expires. Users are paged over personally constructed paging areas. The size of these areas is an increasing function of the time since last user-network contact. The shape of the paging area depends on the particular probability distribution of user location. The problem formulation is applicable to arbitrary motion models assuming the associated user location distribution is available or can be estimated. As an illustrative example, we use a Brownian motion process for user motion and assume Poisson call arrivals. We then investigate the influence of user mobility and call parameters on the optimum timer value and on the optimum configuration of the paging area with P/sub c/ probability of paging success. We show that this scheme can reduce the cost by up to 50% as compared to the conventional timer-based scheme and the cost lower-bounds that of the adaptive location area tracking scheme with a proper choice of the probability criterion P/sub c/.