Detection Performance with Many Antennas Available for Bandwidth-Efficient Uplink Transmission in MU-MIMO Systems

TL;DR

This paper evaluates detection techniques in MU-MIMO uplink systems with many antennas, showing that simple linear detectors combined with interference cancellation can approach optimal performance when the base station has at least five times more antennas than users.

Contribution

It demonstrates that iterative detection with interference cancellation significantly improves performance in large-antenna MU-MIMO uplink systems using simple linear detectors.

Findings

Simple linear detectors alone have high error floors.

Iterative detection with interference cancellation approaches optimal performance.

At least five times more BS antennas than user antennas are needed for near-optimal results.

Abstract

This paper is concerned with SC/FDE for bandwidth-efficient uplink block transmission, with QAM schemes, in a MU MIMO system. The number of BS receiver antennas is assumed to be large, but not necessarily much larger than the overall number of transmitter antennas jointly using the same time/frequency resource at MT. In this context, we consider several detection techniques and evaluate, in detail, the corresponding detection performances (discussed with the help of selected performance bounds), for a range of values regarding the number of available BS receiver antennas. From our performance results, we conclude that simple linear detection techniques, designed to avoid the need of complex matrix inversions, can lead to unacceptably high error floor levels. However, by combining the use of such simple linear detectors with an appropriate interference cancellation procedure - within an iterative DF technique -, a close approximation to the SIMO MFB performance can be achieved after a few iterations, even for 64-QAM schemes, when the number of BS antennas is, at least, five times higher than the number of antennas which are jointly used at the user terminals.