On Improving Gauss-Seidel Iteration for Signal Detection in Uplink Multiuser Massive MIMO Systems

Abstract

Gauss-Seidel (GS) iteration can be used to approximate the minimum mean-squared error (MMSE) result of signal detection in uplink multiuser massive MIMO systems. In this paper, we first propose to employ decision in GS iteration to enhance the detection performance in various aspects. Specifically, GS iteration with decision (GSID) converges faster than the conventional GS iteration and performs even better than the exact MMSE detection in terms of the resultant error rate. By virtue of the use of decision, interference cancelation can be employed to lower the dimension of the signal model in the detection process, and therefore the computational complexity can be significantly reduced. In addition, grey region is introduced in the signal constellation for decision, which lessens the error-propagation effect and further improves the error-rate performance.