An Enhanced Device Localization Approach Using Mutual Signal Strength in Cellular Networks

This study proposes an accurate and calibration-free mobile device localization algorithm in cellular networks. By exploiting the mutual received signal strength (RSS) between base stations, the proposed algorithm creates an accurate RSS-distance mapping using multivariable regression approaches. Unlike traditional methods, the generated mapping is not fixed but rather dynamic to reflect current environments, thus, better characterizing the RSS-distance relationship in the target area and achieving more accurate location estimations without additional calibration effort. Furthermore, this study adopts singular value decomposition to enhance the robustness of our system. Experiments were conducted in a realistic GSM network. Results demonstrated that our approach apparently improves the positioning accuracy, as compared to three existing cellular-based methods, Cell-ID, enhanced Cell-ID, and propagation-model approaches.