Integrated User Grouping, Subcarrier Allocation, and Hybrid Beamforming for Wideband Multiuser mmWave Massive MIMO

Hybrid beamforming structure is widely used in mmWave massive MIMO due to the low hardware complexity. However, its performance is severely degraded by beam squint effects in wideband multiuser mmWave massive MIMO resulting from the non-frequency-specific analog beamforming. To overcome this issue, we develop an integrated user grouping, subcarrier allocation, and hybrid beamforming (IUSH) scheme to compensate for the beam squint effects and maximize the multiuser sum-rate. First, we investigate the time-division and frequency-division analog beamforming for the single radio frequency (RF) chain cases to gain useful insights regarding user grouping and subcarrier allocation. The frequency-division approach achieves superior performance by leveraging the correlations between analog beamforming and user channels, and utilizing subcarrier allocation to dynamically align them. To this end, for multiple RF chain cases, we propose a wideband user grouping strategy that clusters users with close physical channel angles to strengthen the correlations, thereby facilitating efficient subcarrier allocation among users. Then, we develop an alternating minimization-based joint subcarrier allocation and hybrid beamforming (AM-JSH) algorithm to maximize the sum-rate of grouped users. The AM-JSH algorithm alternates between subcarrier allocation and hybrid beamforming, where the subcarrier allocation is formulated as an assignment problem to be solved using the classic Hungarian algorithm, and the hybrid beamforming is implemented with the developed penalty-based two-loop hybrid beamforming algorithm. Simulation results show that the developed IUSH scheme significantly outperforms the existing ones.