Joint maximum likelihood and QRD-M detection algorithm for MIMO system

Maximum likelihood detection (MLD) algorithm provides the best bit error rate (BER) performance for multiple-input multiple-output (MIMO) system. However, the computational complexity of ML detection algorithm grows exponentially with the number of transmit antennas, which results in impractical to use in practice system. The use of QR decomposition with an M-algorithm (QRD-M) has been proposed to provide near ML performance and low complexity. The QRD-M algorithm reduces the complexity by selecting M candidates with the smallest accumulated metrics at each level of the tree search. The trade-off between performance and complexity can be adjusted by setting the parameter M which cannot provide more valuable tradeoff options with better performance to complexity ratio. A new detection scheme, jointing MLD with QRD-M detection algorithm, is proposed in the paper. After performing QR decomposition of the channel matrix, the MLD with length L is done, the accumulated metrics are calculated and sorted, which gives an ordered set, then QRD-M algorithm are used to search the left layers. The proposed algorithm provides better tradeoff options by selecting different parameters L and M and getting more near-ML performance with lower complexity.