Part two: Kalman filtering options for error minimization in statistical terminal assisted mobile positioning

Kalman filtering is a commonly used method for optimizing the performance of techniques that track the mobile terminals’ position given a series of network measurements of Received Signal Strength (RSS). In this paper, we present an analysis of the efficiency of alternative Kalman filter schemes applied to Statistical Terminal Assisted Mobile Positioning (STAMP). In particular, three different options are considered: filtering of the series of RSS measurements, filtering of the series of estimated Mobile Terminal (MT)–Base Station (BS) distances and filtering of the series of estimated MT positions. A two step analysis of the Kalman filter efficiency is conducted: (a) ideal filter performance based on the actual covariance of the input parameters and (b) the actual filter performance based on estimation of the covariance of the input parameters. The analysis is further extended by considering systematic errors inherent in the position estimation process. The paper proves that RSS Kalman filtering is the most efficient method and highlights the sensitivity of the method in systematic errors.