Location Updates In Cellular Networks Using Bloom Filters

Abstract

Location updates (LU) are emergency 911 procedures mandated by the FCC for cellular networks today; helping locate mobiles within 100 meters of their vicinity. This requires paging all mobiles within a vicinity regularly thereby leading to an increased use of bandwidth. This paper analyzes the existing schemes of hash based paging in LU procedures using Bloom filters (BF) and introduces two new schemes to improve performance: optimization of Bloom filters (OBF) and cumulative Bloom filters (CBF). An identifier field in the paging message is coded by applying hashing functions to create a BF and this is used to page a number of mobiles concurrently. False LU are the mobiles that may not belong to a particular paging area but still respond with LU updates. We observe that these false probabilities are very small and can be traded-off with the bandwidth gain. The results obtained compare the analytical and simulation results and their observation leads us to the goal of this research: to obtain a multi-fold increase in bandwidth gain at the cost of keeping the false positives to a realistic minimum