Multi-IRS-Assisted Wireless Localization

Abstract

Intelligent reflecting surface (IRS) is a promising technology that can dynamically control the wireless signal’s amplitude and/or phase, reconfiguring the radio propagation environment. In this letter, we investigate a wireless localization system assisted by multiple IRSs, where virtual line-of-sight (LoS) links are established for aiding an access point (AP) in accurately determining the user’s location. Specifically, the received signals at the AP are first separated into the received signals from different paths by designing specialized phases of IRS. Then, a low-complexity method is employed to estimate the time of arrival (ToA) of each path, and the accuracy of ToA estimation is further improved through the beam training method. Based on the estimation of ToAs, a positioning method is applied to estimate user position. Finally, we derive Cramér-Rao bound (CRB) of all the ToA parameters for user localization and the position error bound (PEB) of user position. Simulation results show that the estimation accuracy of ToA approaches the CRB, and the position approaches the PEB.