Energy Efficient Multi-User Millimeter Wave MIMO Systems Using Tree Search Hybrid Precoding and Few-Bit ADCs

Abstract

Millimeter wave (mmWave) communication is an emerging technique to provide large bandwidth in cellular systems. However, mmWave signals experience more pathloss than the microwave signals adopted in current wireless communications. Therefore, a mmWave system tends to adopt a large number of antennas to combat pathloss with high beamforming gain. In conventional multi-antenna systems, both beamforming and precoding are done digitally in baseband. The high cost and power consumption of certain components in mmWave systems render conventional architecture infeasible. In this paper, we consider low complexity hybrid analog/digital precoding and low-resolution analog-digital converters (ADCs) (1-3 bits) in downlink multiuser mmWave MIMO systems. We propose a new transmission scheme involving rationale of beamspace MIMO communications and beam selection. We also develop a low complexity beam selection algorithm and evaluate adaptive selection on the RF configuration for energy efficiency (EE) optimization. Our analysis and simulation results show that the proposed scheme can significantly reduce system complexity and achieve good performance on energy efficiency and detection.