Adaptive parallel and iterative QRDM detection algorithms for MIMO multiplexing systems

QR-decomposition with M-algorithm (QRDM)achieves quasi-ML performance in multiple-input multiple-output (MIMO) multiplexing systems. Nevertheless, QRDM performs avoidable computations because of its systematic search strategy and its failure to consider the channel and noise conditions. Another drawback is that QRDM is sequential, which limits pipelining capabilities. In this paper, we propose quasi-ML adaptive parallel QRDM (APQRDM) and adaptive iterative QRDM (AIQRDM) algorithms based on set grouping. In set grouping, the tree-search stage of theQRDMalgorithm is divided into partial detection phases (PDPs). These are processed in parallel and iteratively in the proposed APQRDM and AIQRDM algorithms, respectively. Therefore, when the tree-search stage of the QRDM algorithm is divided into G PDPs, the latency of the proposed APQRDM algorithm and the hardware requirements of the proposed AIQRDM algorithm are reduced by a factor of G compared to the QRDM algorithm. Moreover, simulation results show that in 4×4MIMO system, and at E _b/N ₀ of 12 dB, APQRDM decreases the average computational complexity to approximately 43% that of the conventional QRDM. Also, at E _b/N ₀ of 0 dB, AIQRDM algorithm reduces the computational complexity to about 54%, and the average number ofmetric comparisons to approximately 10%, compared to the conventional QRDM.