Near-optimal and efficient MIMO detectors for 64-QAM symbols

This paper presents a simulation study that evaluated one optimal and several near-optimal algorithms for detecting 16-QAM and 64-QAM symbols in single-user MIMO receivers. The study is part of a larger project to develop efficient and near-optimal MIMO detection algorithms for high-order symbol constellations. Fouladi Fard´s near-optimal detection algorithm differs from the conventional V-BLAST algorithm by considering each possible value of the noisiest symbol in the current symbol vector, and then performing V-BLAST in parallel for each candidate to detect the remaining symbols. We investigated modifications of Fouladi Fard´s algorithm that restrict the search space for the first (noisiest) symbol, thereby reducing the computational complexity while still producing performance that significantly exceeds that of V-BLAST.