Joint Hybrid Beamforming Design for IRS-Assisted mmWave Full-Duplex MIMO Systems

This paper proposes a joint hybrid beamforming design for intelligent reflecting surface assisted (IRS-JHBF) millimeter wave (mmWave) full-duplex (FD) multiple-input multipleoutput (MIMO) systems. The proposed IRS-JHBF consists of analog beamformers, digital precoder/combiner and IRS. The analog beamformers are designed by angle-of-departure (AoD) and angle-of arrival (AoA) information of mmWave channel. Then, we employ singular value decomposition (SVD) and minimizing the mean square error (MMSE) to complete the digital precoder and combiner. Finally, the IRS phase shifts are optimized via the multivariate particle swarm optimization (MV-PSO) algorithm to maximize the system sum rate. Simulation results demonstrate that the proposed IRS-JHBF system significantly eliminates the self-interference and increases the achievable sum rate.