A genetic algorithm approach to user location estimation in UMTS networks

Abstract

An innovative solution to user location estimation in 3G networks is presented. The proposed solution uses network information such as Cell-ID and transmitted signal strength, which are easily captured by current 3G mobile handsets, as input to a location detection algorithm which searches the whole network coverage area for the most probable point of origin. This search is performed within a software model of the radio network, which faithfully replicates the provider's live network. The algorithm was optimized through the application of a genetic algorithm to provide an estimate of the user's location in quasi real-time. Simulation results demonstrate that allowing for a 450 m error, the algorithm is capable of locating the user's position in 72% of the cases within an urban environment.