Constellation constrained multiuser multiple‐input multiple‐output for high capacity and error performance over correlated fading channels

Channels' correlation has direct impact to degrade the capacity and reliability of multiple‐input multiple‐output (MIMO) systems considerably. In this paper, new signal constellation designs are investigated to mitigate fading correlation and maximize the capacity and error performance of multiuser MIMO (MU‐MIMO) over correlated channels, which is a major research challenge. Two methods are studied in a novel constellation constrained MU‐MIMO approach, namely, unequal power allocation and rotated constellation. Based on principles of maximizing the minimum Euclidean distance (d_min) of composite received signals, users' data can be recovered using maximum likelihood joint detection irrespective of correlation values. Compared with the identical constellation scenario in conventional MU‐MIMO, it is shown that constellation rearrangement of transmitted signals has direct impact to resolve the detection ambiguity when the channel difference is not sufficient, particularly in moderate to high correlations. Extensive analysis and simulation results demonstrate the superiority of proposed technique to capture most of the promised gains of multiantenna systems and application for future wireless communications. Copyright © 2014 John Wiley & Sons, Ltd.