Hybrid zero‐forcing for correlated and semi‐orthogonal multi‐user MU‐MIMO channels

Abstract

Fifth generation cellular networks must support dense user environments where close proximity users and highly correlated channels are prevalent. This study presents a novel decorrelating zero-forcing (DZF) scheme. It improves downlink zero-forcing (ZF) performance over multi-user multiple-input multiple-output channels when highly correlated users are present. DZF groups correlated users into pairs where the stronger user is labelled ‘near’ and the weaker user, ‘far’. For the far and unpaired users, the proposed design is similar to conventional ZF (CZF) in that the base station performs traditional ZF and the users employ the leading left singular vectors of the channel for processing. In contrast, the near users employ the second strongest left singular vector which decorrelates their effective channel with the paired far user. This reduces the noise inflation inherent in ZF for correlated channels. From simulations and analysis the authors demonstrate that DZF has both rate and fairness advantages over CZF when highly correlated users are present. Furthermore, they design a hybrid ZF (HZF) scheme, which harnesses the advantages of both CZF and DZF, hence providing robustness against the joint scheduling of semi-orthogonal and highly correlated users with very little additional complexity. Simulations clearly demonstrate that HZF has higher rates than CZF and DZF.