Geometric Location Techniques With SSSD in Wireless Cellular Networks: A Comparative Performance Study

Geometric location techniques involve using either distance or distance difference, which often results in a nonlinear least square (NLLS) estimation. Although many studies to tackle the NLLS estimation by iterative algorithm have been proposed, however in a practical environment for both distance and distance difference location techniques, there is no reported work on the achievable performance and still lacks a comparative performance study. Based on the stationary signal-strength- difference (SSSD) measurement data obtained from an experiment in a commercial wireless cellular network, we examine and compare the performance of both distance and distangradientce difference location techniques, using two iterative algorithms, namely, Levenberg-Marquardt algorithm (LMA) and gradient descent algorithm (GDA). Results demonstrate that for both types of location techniques the performance achieved by LMA is significantly better than that achieved by GDA, and for both types of algorithms the performance of distance location technique is better than that of distance difference location technique.