Low Complexity Joint User Scheduling and Hybrid Beamforming for mmWave Massive MIMO Systems

Codebook-based hybrid analog and digital beam-forming (HBF) plays a key role in practical millimeter wave (mmWave) massive MIMO systems. Most existing work focuses on the design of analog beams, without considering the coupled relationship between beam selection and user scheduling. In this paper, we consider a downlink HBF system and address the aforementioned problem. By exploiting the orthogonality of the Discrete Fourier Transform (DFT) codebook, we propose two low complexity approaches named as E-DFT and PF-DFT according to different requirements. Specifically, E-DFT aims to maximize the sum achievable rate of the HBF system, and PF-DFT considers both the system throughput and user fairness. Simulation results show that our proposed E-DFT scheme outperforms the baseline methods in the literature, and achieves near optimal sum rate. Meanwhile, the PF-DFT scheme performs better in balancing the achievable rate and scheduling fairness.