Ordered multi-branch processing for successive interference cancellation based MIMO detection

Abstract

In this article, we propose an ordered multi-branch (OMB) processing for successive interference cancellation (SIC) based symbol vector detection techniques in multiple-input multiple-output (MIMO) spatial multiplexing systems. In the proposed method multiple ordered branches are initiated in parallel where each branch employ SIC based detection with different ordering pattern. To generate multiple branches, in the proposed work, first we consider channel norm (ChN) and log likelihood ratio (LLR) based orderings of the detection sequence, and then difference in the values of ChN and/or LLR of consecutive layers are used. Each branch produces an estimate of the transmitted symbol vector and the best one is selected by using the maximum likelihood (ML) rule. To improve the accuracy of decisions in each branch and achieve a better performance, we also incorporate multiple feedback (MF) strategy based SIC algorithm for symbol vector detection in MIMO systems. Since, all branches process the input vector independently, the proposed algorithm is well suited (in terms of processing speed) for parallel processing receiver architectures. The bit error rate (BER) performance and the computational complexity of the proposed technique is computed, and compared with SIC and multiple feedback SIC (MF-SIC) based MIMO detection algorithms. Simulation results reveal that the proposed methods outperform SIC and MF-SIC algorithms, and achieve near ML performance with less computational complexity.