A Kalman-Filter Based Paging Strategy for Cellular Networks

Developing an efficient location management technique is an important step in working towards the determination of an optimal solution to the problem of managing mobility. Given the irregular nature of cell sizes in a cellular network, the behavior of mobile movement changes from cell to cell and from user to user. Thus, the need for designing an adaptive algorithm for tracking a roaming mobile becomes imperative. In this paper, we propose a new predictive location management strategy that reduces the update cost while restricting the paging load optimized for mobiles roaming with traceable patterns. Enhanced with directional predictive capabilities offered by Kalman filtering, new update boundaries are assigned to better reflect the movement patterns of individual mobiles upon location registration. Thus, while complying with the required delay constraints, QoS measures (mainly throughput) will not need to be sacrificed as a result of increasing the update threshold. The contribution of this paper is two-fold: (1) to suggest a distribution model that is capable of describing a wide range of movement patterns with varying correlation between traveling directions and (2) to explore the capabilities in terms of reliable performances of Kalman filter in predicting future movement patterns. Simulation results have successfully demonstrated the ability of Kalman filter in assigning update boundaries capable of reflecting a mobile's roaming characteristics. The performance gains, achieved mainly through a significant reduction in the number of updates, indicate its potential for promoting better bandwidth conservation.