Optimization of Precoder and Combiner in mmWave Hybrid Beamforming Systems for Multi-user Downlink Scenario

Beamforming at millimeter wave (mmWave) band, promises to significantly support 5G networks in achieving their performance goals. The conventional digital beamforming uses a separate RF chain for each antenna element, while it leads to high cost and hardware complexity in mmWave massive MIMO antenna systems. Beamforming with multiple data streams called precoding improves the system's spectral efficiency and one of its kind hybrid beamforming reduces the cost and overcomes the hardware limitation by using reduced number of RF chains. This work considers, transmit precoding, receive combining in mmWave hybrid beamforming systems and constructs a dictionary matrix containing array response vectors. This paper proposes an extended simultaneous orthogonal matching pursuit (ESOMP) algorithm to compute the block-sparse matrix. The non-zero rows of block-sparse matrix and dictionary matrix are further processed to achieve precoder/combiner optimization in multi-user downlink scenario. Simulation results reveal that the proposed method performs close to the ideal digital beamforming scheme while improving the spectral efficiency when compared to the state-of-the-art algorithm.