Performance Analysis of Hybrid Beamforming Techniques in Large MU MIMO Systems

Abstract

AbstractMassive multiple input multiple output (MIMO) at mmWave frequency is one of the primary suggestions for next-generation networks in order to support this large mobile data load. The large number of antennas at the base station (BS) allows multiple users (MU) to share the same time-frequency resources and concentrate energy in a smaller area. As a result, massive MIMO can significantly enhance spectrum and energy efficiency. However, because of the needed number of RF chains per antenna element, fully digital beamforming is undesirable when the number of antennas at the BS becomes very large, resulting in excessive complexity and high power consumption. In this work, we evaluate the performances of hybrid beamforming techniques in downlink MU mmWave large MIMO systems. Specifically, we consider, hierarchical singular value decomposition (SVD) and two-stage iterative MU hybrid beamforming algorithms to alleviate the optimization complexity on finding optimal analog and digital precoding matrices. First, based on the mmWave system and channel model, we formulate the spectral efficiency expression and theoretical analysis is done. Then, the proposed hybrid beamforming algorithms are formulated and numerical simulation are done to evaluate the spectral efficiency and computational complexity of the algorithms. The results reveal that fully-connected hybrid beamforming techniques outperform partially-connected hybrid beamforming techniques in terms of spectrum efficiency.KeywordsEnergy efficiencyHybrid beamforming techniquesLarge MIMOmmWaveSpectral efficiency