Cramér-Rao lower bound analysis for wireless localization systems using priori information

Abstract

Bayesian estimation methods are widely used for wireless localization systems. They employ priori information and current measurement error distribution models to derive the state of a mobile target. Cramér-Rao lower bound (CRLB) is a fundamental tool to analyze the performance of Bayesian estimators. Although CRLB is derived based on the measurement error distribution, only a few works have investigated the performance using priori information. In this paper, we derive the CRLB formulation in three cases by using the priori information: (1) fundamental Bayesian process; (2) recursive process; (3) adaptive process. These three processes represent the common Bayesian tracking algorithms for wireless system. Simulations are constructed to compare the localization performance according to the different processes. The results indicate how the priori information influences the location estimation and how to improve the performance according to different scenarios.