Near-Optimal Hybrid Analog and Digital Precoding for Downlink mmWave Massive MIMO Systems

TL;DR

This paper introduces a low-complexity, near-optimal hybrid precoding scheme for mmWave massive MIMO systems that improves performance by considering realistic subarray structures and iterative optimization.

Contribution

It proposes an iterative hybrid precoding method based on subarray structures, reducing complexity and enhancing performance over existing schemes.

Findings

Achieves better spectral efficiency than recent methods

Maintains acceptable computational complexity

Effectively utilizes successive interference cancellation

Abstract

Millimeter wave (mmWave) massive MIMO can achieve orders of magnitude increase in spectral and energy efficiency, and it usually exploits the hybrid analog and digital precoding to overcome the serious signal attenuation induced by mmWave frequencies. However, most of hybrid precoding schemes focus on the full-array structure, which involves a high complexity. In this paper, we propose a near-optimal iterative hybrid precoding scheme based on the more realistic subarray structure with low complexity. We first decompose the complicated capacity optimization problem into a series of ones easier to be handled by considering each antenna array one by one. Then we optimize the achievable capacity of each antenna array from the first one to the last one by utilizing the idea of successive interference cancelation (SIC), which is realized in an iterative procedure that is easy to be parallelized. It is shown that the proposed hybrid precoding scheme can achieve better performance than other recently proposed hybrid precoding schemes, while it also enjoys an acceptable computational complexity.