Mobile Station Positioning Using GSM Cellular Phone and Artificial Neural Networks

Abstract

In this paper, we describe a novel approach to mobile station positioning using a GSM mobile phone. The approach is based on the use of an inherent feature of the GSM cellular system (the mobile phone continuously measures radio signal strengths from a number of the nearest base stations (antennas)) and on the use of this information to estimate the phone's location. The current values of the signal strengths are processed by a trained artificial neural network executed at the computer attached to the mobile phone to estimate the position of the mobile station in real time. The neural network configuration is obtained by using a genetic algorithm that searches the space of specific neural network types and determines which one provides the best location estimation results. Two general methods are explored: the first is based on using a neural network for classification and the second uses function approximation. The experimental results are reported and discussed.