Location estimation using RSS measurements with unknown path loss exponents

Abstract

In cellular networks, the location of an MS can be estimated using received signal strength (RSS) measurements from the control channels of several base stations. RSS-based location estimation techniques provide a low-cost, low-complexity solution for LBS. However, the positioning accuracy of these techniques is often inadequate for many LBS. In most of the recent studies, signal propagation characteristics are assumed to be stable and known. Such an assumption causes a degradation on the positioning accuracy in many practical application scenarios. In this thesis, an RSS-based location estimation technique, so called RSS-MPLE, that jointly estimates the propagation parameters and the MS position is proposed. RSS-MPLE method incorporates the antenna radiation pattern information into the signal model and finds the Maximum Likelihood (ML) estimate of unknown parameters by employing Levenberg-Marquardt method, which is a nonlinear least-squares minimization algorithm. Performance of RSS-MPLE algorithm is evaluated under various scenarios and simulation results show that RSS-MPLE algorithm offers an accurate position estimator that is robust against the variations in the signal propagation characteristics.