A New Candidate Adding Algorithm for Coded MIMO Systems with Fixed-Complexity Detection

We propose a new approach that achieves near max-log optimal performance by improving soft information of the eliminated bits in reduced and fixed complexity coded multiple-input multiple-output (MIMO) systems. As a fixed-complexity MIMO detection, we consider both the QR decomposition with M-algorithm (QRM) and the fixed-complexity sphere decoding (FSD). In general, for the MIMO system with channel coding, the detector needs to calculate soft information for each coded bit. However, the reduced complexity detection often eliminates the candidate symbols that are required for accurate soft information calculation. We propose a novel symbol recovery approach that can retrieve the coded bits and thus recalculate the soft information with improved error performance. Simulation results demonstrate the effectiveness of the proposed soft information calculation approach under the example system with 4 x 4 MIMO-OFDM employing 16-QAM and 64-QAM.