Iterative QR decomposition-based detection algorithms with multiple feedback and dynamic tree search for LDPC-coded MIMO systems

In this paper, we present two innovative LDPC- coded QR decomposition-based soft-output detection techniques, both of which are able to achieve a near-ML performance with significant reduced complexity compare to other optimal detection solutions, such as MAP or list SD algorithms. The first detector (MF-QRD) employs a multi-feedback technique to select appropriate candidates when the symbols are unreliable. Another detection strategy called variable-M QRD (VM-QRD) detector is developed which dynamically adapts the number of detection candidates according to the channel variations in each detection layer. The irregular PEG LDPC code is employed as the outer channel code which provides efficient redundancy for mitigating remaining co-channel interference and additive noise. And simulation results show that the proposed algorithms have excellent performances.