Kalman-filter-based predictive location management for PCS networks

Location management in personal communication services (PCS) networks in general consists of two processes: paging and location updating. A mobile user is registered to a location area (LA), which consists of several neighboring cells. When an incoming call arrives, the system pages the corresponding mobile user by sending a paging message in every cell within his registered LA. We present in this paper a predictive location management scheme that uses Kalman filter to estimate the velocity and predict the future movement of a mobile user. Based on the predicted location and the estimated variance and covariance of the x- and y-coordinates, we move the LA in the direction of the estimated velocity, and we also change its shape dynamically. This method is adaptive to the mobility pattern of each mobile user. The performance is obtained by simulation and compared to the static and movement-based location management schemes.