Singular Value Thresholding based Adaptive Approach for Hybrid Beamforming in MmWave Massive MIMO-OFDM Transmitters

Thanks to the high spectral efficiency and the abundant bandwidths, millimeter-wave (mmWave) communication combined with massive multiple-input multiple-output (MIMO) technology have been considered as a promising candidate for next generation mobile communication systems. As a mmWave massive MIMO system requires a huge number of radio frequency (RF) chains, hybrid analog/digital (A/D) beamforming (BF) is addressed as the best solution to surpass digital BF limitations and reach lower cost and power consumption. Therefore, new specific A/D BF algorithms for mmWave massive MIMO systems are required. This paper considers the problem of BF design in a hybrid mmWave massive MIMO-OFDM system using the sub-array structure which reduces the number of phase shifters since each RF chain should be only connected to a specific sub-array. For this structure, we propose an adaptive hybrid Am BF design based on singular value thresholding algorithm. Then, our proposal is evaluated in terms of spectral efficiency, outperforming the semidefinite relaxation (SDR) based algorithm and providing an important reduction in terms of computational complexity.