RIS-Assisted High Resolution Radar Sensing

TL;DR

This paper demonstrates how reconfigurable intelligent surfaces (RIS) can significantly enhance high-resolution radar sensing by improving detection and localization accuracy, especially in scenarios with unresolved multiple targets.

Contribution

It introduces a RIS beam sweep procedure, derives bounds for channel estimation and detection, and proposes an OMP-based algorithm for improved sensing accuracy.

Findings

RIS improves detection probability and localization accuracy

Proposed algorithms effectively fuse RIS and non-RIS signals

Numerical results confirm RIS's potential for high-resolution sensing

Abstract

This paper analyzes monostatic sensing by a user equipment (UE) for a setting in which the UE is unable to resolve multiple targets due to their interference within a single resolution bin. It is shown how sensing accuracy, in terms of both detection rate and localization accuracy, can be boosted by a reconfigurable intelligent surface (RIS), which can be advantageously used to provide signal diversity and aid in resolving the targets. Specifically, assuming prior information on the presence of a cluster of targets, a RIS beam sweep procedure is used to facilitate the high resolution sensing. We derive the Cram\'er-Rao lower bounds (CRLBs) for channel parameter estimation and sensing and an upper bound on the detection probability. The concept of coherence is defined and analyzed theoretically. Then, we propose an orthogonal matching pursuit (OMP) channel estimation algorithm combined with data association to fuse the information of the non-RIS signal and the RIS signal and perform sensing. Finally, we provide numerical results to verify the potential of RIS for improving sensor resolution, and to demonstrate that the proposed methods can realize this potential for RIS-assisted high resolution sensing.