A Dual-Function Radar-Communication System Empowered by Beyond Diagonal Reconfigurable Intelligent Surface

TL;DR

This paper introduces a novel beyond diagonal reconfigurable intelligent surface (BD-RIS) for dual-function radar-communication systems, enabling full-space coverage and improved performance through joint optimization of system components.

Contribution

It proposes a new BD-RIS architecture supporting hybrid reflection and transmission, and develops an optimization algorithm for enhanced radar and communication functions.

Findings

BD-RIS achieves full-space coverage.

Proposed algorithm improves target detection and communication quality.

System outperforms conventional RIS in simulations.

Abstract

This work focuses on the use of reconfigurable intelligent surface (RIS) in dual-function radar-communication (DFRC) systems to improve communication capacity and sensing precision, and enhance coverage for both functions. In contrast to most of the existing RIS aided DFRC works where the RIS is modeled as a diagonal phase shift matrix and can only reflect signals to half space, we propose a novel beyond diagonal RIS (BD-RIS) aided DFRC system. Specifically, the proposed BD-RIS supports the hybrid reflecting and transmitting mode, and is compatible with flexible architectures, enabling the system to realize full-space coverage and to achieve enhanced performance. To achieve the expected benefits, we jointly optimize the transmit waveform, the BD-RIS matrices, and sensing receive filters, by maximizing the minimum signal-to-clutter-plus-noise ratio for fair target detection, subject to the constraints of the communication quality of service, different BD-RIS architectures and power budget. To solve the non-convex and non-smooth max-min problem, a general solution based on the alternating direction method of multipliers is provided. Numerical simulations validate the efficacy of the proposed algorithm and show the superiority of the BD-RIS aided DFRC system in terms of both communication and sensing compared to conventional RIS aided DFRC.