Simultaneous Active-Passive Beamformer Design in IRS-Enabled Multi-Carrier DFRC System

Intelligent reflecting surfaces (IRS) are increasingly considered as an enabling technology for non-line-of-sight (NLoS) communications and remote sensing. We consider an IRS-aided dual-function radar-communications (DFRC) system for multi-carrier waveforms in a combined light-of-sight and NLoS scenario. In this wideband system, we jointly design the radar receive filter, frequency-dependent beamforming, and IRS phases to maximize the average radar signal-to-interference-plus-noise ratio (SINR) and the minimal communications SINR among all users. The problem has a maximin objective function with constant modulus and power constraints and is, therefore, highly non-convex. We solve this optimization using the alternating maximization framework, which iterates amongst the alternating direction method of multipliers, Riemannian steepest descent, and Dinkelbach's algorithms. Numerical experiments demonstrate that the proposed method for IRS-aided DFRC achieves better performance for both sensing and communications than the system without IRS.