Enabling Coherent Multipath Estimation in RIS-Assisted Systems via a Synthesized Virtual Array Design

Reconfigurable intelligent surfaces (RISs) have emerged as a promising technology for enabling high-precision wireless sensing. However, existing angular estimation methods for RIS-assisted systems face challenges in coherent multipath environments due to the rank deficiency of signal covariance matrices. To address this, this paper proposes a time-domain synthesized virtual array framework through RIS configuration design. Specifically, we design a sliding-window reflection protocol that reorganizes temporal signal samples into a virtual uniform linear array structure, effectively embedding spatial information into the time domain. We theoretically prove that the virtual structure enables the application of forward-backward spatial smoothing (FBSS), guaranteeing full-rank covariance restoration for robust angular estimation. Simulation results validate the resolution of coherent multipath and reveal that prioritizing virtual aperture over snapshot diversity maximizes precision under constrained resources.