Localization Based on Narrowband Power-Only Measurements Using Distributed Reconfigurable Intelligent Surfaces: Concept and Experimental Validations

Localization is expected to support advanced location-aware services in the beyond fifth/sixth generation (B5G/6G) wireless networks. Conventional methods typically rely on wideband or multi-antenna systems to extract delay or spatial features, which are infeasible for low-complexity, narrowband devices that report only received power. The advent of reconfigurable intelligent surfaces (RISs) enables the intelligent yet energy- and hardware-efficient reflection of the feeding signal with controllable reflection phase shifts via software, making high-accuracy positioning and extended service coverage possible. Unlike existing RIS-based localization methods relying on wideband complex signals, this paper introduces a novel two-stage method for narrowband power-only localization leveraging two or more distributed RISs. The first stage comprises the initial Tx-RIS-Rx response estimation via power-only fast calibration based on the rotating element electric field vector (REV) method, followed by the RIS-Rx channel response reconstruction. In the second stage, with the obtained RIS-Rx channel response, the multiple signal classification (MUSIC)-based estimation can be performed to first estimate the angle of departures (AoDs) with respect to multi-distributed RISs, and then the Rx position is estimated via intersection search. The estimation root mean square error (RMSE) versus received signal-to-noise ratio (SNR) was simulated to demonstrate the accuracy and effectiveness of the proposed algorithm in practical low-SNR environments. Benchmarked by the existing RIS beam-searching method, the proposed localization method is experimentally validated for 12 Rx positions by utilizing a 2-bit RIS-assisted measurement system. The measurement results demonstrate that the proposed method offers high-precision localization in narrowband power-only regimes, making it preferred in scenarios that require a fast response and low computational load compared with the conventional wideband complex-signal-based methods.