MMSE Hybrid Beamforming for Multi-User Millimeter Wave MIMO Systems
Wenqian Ren, Jin Deng, Xiantao Cheng
Abstract
Millimeter wave (mmWave) massive multiple-input multiple-output (MIMO) system has been recognized as a key technique for 5G communications. In terms of power consumption and hardware cost, it is hardly affordable to assign one radio frequency (RF) chain to each antenna in massive MIMO. Therefore, hybrid analog and digital beamforming is often adopted to reduce the number of RF chains. In this letter, we address the hybrid precoding and combining for multi-user mmWave MIMO systems. The goal is to minimize the bit error rate (BER) in downlink data transmission from the base station (BS) to multiple users. To fulfill this, we choose the sum of the users’ mean square errors (sum-MSE) as the optimization objective, and develop an iterative algorithm to design the BS precoding matrices and the combining matrices of each user by gradually minimizing the sum-MSE. Compared with the current hybrid beamforming schemes, the proposed scheme can achieve much better BER performance.