Detection-Aided Ordering for LMMSE-ISIC in MIMO Systems

In this paper, the detection-aided ordering schemes are proposed for linear minimum mean-squared-error (LMMSE) iterative soft interference cancellation (ISIC) in multiple-input multiple-output (MIMO) systems. Unlike the conventional LMMSE-ISIC ordering schemes that utilize the channel state information (CSI) only, the proposed ordering schemes utilize the receive signal vector and CSI for the ordering procedure. Then, for each candidate symbol, the sum of the likelihoods except the largest likelihood is calculated to estimate the reliability of the candidate symbol, where the likelihoods are calculated by the LMMSE or LMMSE-ISIC detection-aided ordering procedure. Thus, the proposed ordering schemes can provide a significantly more accurate ordering result than the conventional ordering schemes. As the detection-aided ordering schemes, non-iterative and iterative ordering schemes are proposed, and the constrained iterative ordering scheme is also proposed to resolve the high computational complexity of the original iterative ordering scheme. Numerical simulation results verify that the proposed detection-aided ordering schemes outperform the conventional ordering schemes in terms of convergence speed and error performance.