Minorization-based Low-Complexity Design for IRS-Aided ISAC Systems

TL;DR

This paper introduces a low-complexity method for designing IRS-aided ISAC systems by employing a double minorization technique to simplify the non-convex optimization problem, leading to efficient IRS parameter computation.

Contribution

The paper proposes a novel double minorization approach to reduce the complexity of IRS design in ISAC systems with non-convex constraints.

Findings

The method achieves effective IRS parameter design with reduced computational complexity.

Numerical results demonstrate the proposed approach's effectiveness in improving system performance.

The approach simplifies a challenging fourth-order optimization problem into a tractable linear form.

Abstract

A low-complexity design is proposed for an integrated sensing and communication (ISAC) system aided by an intelligent reflecting surface (IRS). The radar precoder and IRS parameter are computed alternatingly to maximize the weighted sum signal-to-noise ratio (SNR) at the radar and communication receivers. The IRS design problem has an objective function of fourth order in the IRS parameter matrix, and is subject to highly non-convex unit modulus constraints. To address this challenging problem and obtain a low-complexity solution, we employ a minorization technique twice; the original fourth order objective is first surrogated with a quadratic one via minorization, and is then minorized again to a linear one. This leads to a closed form solution for the IRS parameter in each iteration, thus reducing the IRS design complexity. Numerical results are presented to show the effectiveness of the proposed method.