A Cramer–Rao Lower Bound of CSI-Based Indoor Localization

Abstract

Due to robustness against multipath effect, frequency domain CSI (Channel State Information) of OFDM (Orthogonal Frequency Division Multiplexing) systems is supposed to provide good ranging measurement for indoor localization. Since lower bound of positioning error for CSI-based localization has been rarely reported, this paper proposes a Cramer-Rao lower bound (CRLB) for CSI-based localization. This CRLB is derived based on indoor wireless propagation model integrating both shadowing effect and multipath effect. The main factors that can influence the CRLB are analyzed. Through simulations and analysis, the gap between CSI-based localization error and the proposed CRLB is presented, and the influence by three main factors is thoroughly investigated.