Intelligent Reflecting Surfaces for Visible Light Positioning Based on Received Power Measurements

In this paper, we formulate and analyze a received power based position estimation problem for visible light positioning (VLP) systems in presence of intelligent reflecting surfaces (IRSs). In the proposed problem formulation, a visible light communication (VLC) receiver collects signals from a number of light emitting diode (LED) transmitters via line-of-sight (LOS) paths and/or via reflections from IRSs. We derive the Cramér--Rao lower bound (CRLB) expression and the maximum likelihood (ML) estimator for generic three-dimensional positioning in the presence of IRSs with arbitrary configurations. In addition, we consider the problem of optimizing the orientations of IRSs when line-of-sight (LOS) paths are blocked, and propose an optimal adjustment approach for maximizing the received powers from IRSs based on analytic expressions, which can be solved in closed form or numerically. Since the optimal IRS orientations depend on the actual position of the VLC receiver, an N-step localization algorithm is proposed to perform adjustment of IRS orientations in the absence of any prior knowledge about the position of the VLC receiver. Performance of the proposed approach is evaluated via simulations and compared against the CRLB. It is deduced that although IRSs do no provide critical improvements in positioning accuracy in the presence of LOS signals from a sufficient number of LED transmitters, they can be very important in achieving accurate positioning when all or most of LOS paths are blocked.