Enhanced RSS Fingerprinting Localization with Reconfigurable Intelligent Surface

Reconfigurable intelligent surface (RIS) holds significant potential to enhance wireless communications by offering good performance, high security, and great efficiency. In particular, RIS provides notable benefits in improving wireless location accuracy for user equipments (UEs), which are often overlooked. In this paper, we utilize RIS for received signal strength (RSS)-based indoor fingerprinting localization. To improve positioning accuracy, we propose a two-stage RIS configuration selection approach. Specifically, in the first stage, we select the RIS configuration that contributes significantly to positioning. In the second stage, a supervised learning approach is used to select features, effectively reducing the large state space of the RIS. The effectiveness of the proposed RIS-assisted RSS fingerprinting localization technique is validated through simulation and field test results.