Efficient Precoding for Single Carrier Modulation in Multi-User Massive MIMO Networks

TL;DR

This paper introduces a regularized zero forcing precoding method for single carrier massive MIMO systems that reduces computational complexity and improves performance at low to moderate SNR levels, validated through analysis and simulation.

Contribution

A novel RZF precoding approach that reuses uplink detection calculations, lowering complexity and enhancing performance over ZF precoding in massive MIMO systems.

Findings

RZF precoding reduces computational complexity compared to ZF.

RZF outperforms ZF at low and moderate SNR levels.

Performance of RZF and ZF converges at high SNR.

Abstract

By processing in the frequency domain (FD), massive MIMO systems can approach the theoretical per-user capacity using a single carrier modulation (SCM) waveform with a cyclic prefix. Minimum mean squared error (MMSE) detection and zero forcing (ZF) precoding have been shown to effectively cancel multi-user interference while compensating for inter-symbol interference. In this paper, we present a modified downlink precoding approach in the FD based on regularized zero forcing (RZF), which reuses the matrix inverses calculated as part of the FD MMSE uplink detection. By reusing these calculations, the computational complexity of the RZF precoder is drastically lowered, compared to the ZF precoder. Introduction of the regularization in RZF leads to a bias in the detected data symbols at the user terminals. We show this bias can be removed by incorporating a scaling factor at the receiver. Furthermore, it is noted that user powers have to be optimized to strike a balance between noise and interference seen at each user terminal. The resulting performance of the RZF precoder exceeds that of the ZF precoder for low and moderate input signal-to-noise ratio (SNR) conditions, and performance is equal for high input SNR. These results are established and confirmed by analysis and simulation.