Joint Design of Hybrid and Reflection Beamforming for RIS-Aided mmWave MIMO Communications

Analog/digital hybrid beamforming architectures have been often considered in millimeter-wave (mmWave) massive multiple-input multiple-output (MIMO) communication systems to address the serious propagation loss problem of mmWave channels. As a consequence, however, the highly directional beam makes mmWave communication very assailable to blockage. Recently emerged reconfigurable intelligent surface (RIS), which can smartly manipulate the propagation environment, is ideal for mmWave MIMO communication to create effective non-line-of-sight (non-LoS) links, bypassing the blockage. In this paper, we consider RIS-aided mmWave MIMO communication systems, where both the transceivers adopt analog/digital hybrid beamforming architectures. For practical application, low-resolution phase shifters (PSs) are used to realize the analog beamforming and the RIS. Aiming at maximizing the spectral efficiency, we propose to jointly design hybrid beamforming and RIS reflection beamforming. To this end, we first decompose the original optimization problem to make it more tractable. Then we utilize an alternating optimization to jointly optimize the analog beamforming and the RIS reflection. Furthermore, the digital beamforming is optimized through typical singular value decomposition (SVD) to further improve the spectral efficiency. Moreover, we analyze and compare the computational complexity of the proposed algorithm and other state-of-the-art algorithms. Simulation results verify the advantages of our proposed joint hybrid beamforming and RIS reflection beamforming design algorithm compared with other algorithms.