A Low-Complexity Design for IRS-Assisted Secure Dual-Function Radar-Communication System

TL;DR

This paper proposes a low-complexity, alternating optimization method for IRS-assisted dual-function radar-communication systems to enhance physical layer security by jointly designing radar waveforms, artificial noise, and IRS parameters.

Contribution

It introduces a novel, low-complexity approach using fractional programming and closed-form solutions for joint design in IRS-assisted secure radar-communication systems.

Findings

Improved secrecy rate compared to baseline methods

Efficient convergence of the proposed optimization algorithm

Enhanced beamforming performance under SNR constraints

Abstract

In dual-function radar-communication (DFRC) systems the probing signal contains information intended for the communication users, which makes that information vulnerable to eavesdropping by the targets. We study the security of a DFRC system aided by an intelligent reflecting surface (IRS) from the physical layer security (PLS) perspective. The IRS helps overcome path loss or blockage and introduces more degrees of freedom for system design, however, it also makes the design problem more challenging. In the system considered, the radar embeds artificial noise (AN) in the probing waveform, and the radar waveform, the AN noise and the IRS parameters are designed to optimize the communication secrecy rate while meeting radar signal-to-noise ratio (SNR) constraints. The contribution of the paper is a novel, low complexity approach to solve the underlying optimization problem and obtain the design parameters. In particular, we consider an alternating optimization approach, where in each iteration, the problem is decomposed into two sub-problems, namely, one that designs the IRS parameters, and another that jointly designs the radar waveform and the AN. The challenges in those sub-problems are the fractional objective, the SNR being a quartic function of the IRS parameters, and the unit-modulus constraint on the IRS parameters. A fractional programming technique is used to transform the fractional form objective into a more tractable non-fractional polynomial form. A closed-form based approach is proposed for the IRS design problem, which results in low complexity IRS design. Numerical results are provided to demonstrate the convergence properties of the proposed system design method, the secrecy rate and beamforming performance of the designed system.