Mobile terminal location estimation using support vector machines

Abstract

Mobile technology advancements and customer demands have pushed mobile operators to develop services which are based on user position. These applications require accurate location estimation which must be achieved through information that is available on the network. The procedure adopted in the proposed solution is based on Cell-ID and received signal strength measurements, which are both available to 3G mobile handsets. The collected information is transmitted to a server and fed to two Support Vector Machines (SVM) connected in cascade to infer location estimation. Simulation results show that the proposed method manages to achieve good accuracies when compared to similar solutions based on received signal strength measurements.