RIS-Aided Bistatic Radar for Rapid NLOS Sensing in the Teraharetz Band

TL;DR

This paper presents a RIS-assisted bistatic radar system operating at terahertz frequencies for rapid NLOS sensing, achieving centimeter to sub-millimeter localization accuracy by combining hierarchical codebooks and iterative maximum likelihood estimation.

Contribution

It introduces a novel RIS-based approach for NLOS sensing at terahertz frequencies, with an iterative MLE scheme that approaches the Cramer-Rao lower bound for localization accuracy.

Findings

Achieves centimeter-level initial localization accuracy.

Refines to sub-millimeter accuracy with additional transmissions.

Effectively localizes targets under NLOS conditions using RIS.

Abstract

In this paper, we investigate a non-lineof-sight (NLOS) sensing problem at terahertz frequencies. To be able to observe the targets shadowed by a blockage, we propose a method using reconfigurable intelligent surfaces (RIS). We employ a bistatic radar system and scan the obstructed area with RIS using hierarchical codebooks (HCB). Moreover, we propose an iterative maximum likelihood estimation (MLE) scheme to yield the optimum sensing accuracy, converging to Cramer-Rao lower bound (CRLB). We take band-specific effects such as diffraction and beam squint into account and show that these effects are relevant factors affecting localization performance in RIS-employed radar setups. The results show that under NLOS conditions, the system can still localize all the targets with very good accuracy using the RIS. The initial estimates obtained by the HCBs can provide centimeter-level accuracy, and when the optimal performance is needed, at the cost of a few extra transmissions, the proposed iterative MLE method improves the accuracy to sub-millimeter accuracy, yielding the position error bound.