RSS Fingerprinting Based Multi-user Outdoor Localization Using Reconfigurable Intelligent Surfaces

Abstract

The received signal strength (RSS) fingerprinting based localization is a widely used technique to locate mobile devices in cellular networks. However, traditional RSS finger-printing based techniques require a set of RSS values from multiple base stations (BSs) at each location, and the RSS from remote BSs is weak and easily affected by the noise and other variations of the propagation channel, which limits the localization accuracy. In this paper, we propose to utilize a reconfigurable intelligent surface (RIS) to improve the accuracy for the RSS fingerprinting based multi-user outdoor localization, which requires the RSS from only one BS. By modifying phase shifts of signals reflected by the RIS, we can create different RSS values at the same location using the signal from the nearest BS for localization. To optimize the RIS phase shifts, we formulate the localization error minimization (LEM) problem and propose an LEM algorithm. Simulation results validate the effectiveness of the proposed scheme.