摘要
Two reduced-complexity decoding algorithms for unitary space-time codes based on tree-structured constellation are presented. In this letter original unitary space-time constellation is divided into several groups. Each one is treated as the leaf nodes set of a subtree. Choosing the unitary signals that represent each group as the roots of these subtrees generates a tree-structured constellation. The proposed tree search decoder decides to which sub tree the receive signal belongs by searching in the set of subtree roots. The final decision is made after a local search in the leaf nodes set of the se-lected sub tree. The adjacent subtree joint decoder performs joint search in the selected sub tree and its “surrounding” subtrees,which improves the Bit Error Rate (BER) performance of purely tree search method. The exhaustively search in the whole constellation is avoided in our proposed decoding al-gorithms,a lower complexity is obtained compared to that of Maximum Likelihood (ML) decoding. Simulation results have also been provided to demonstrate the feasibility of these new methods.
Two reduced-complexity decoding algorithms for unitary space-time codes based on tree-structured constellation are presented. In this letter original unitary space-time constellation is divided into several groups. Each one is treated as the leaf nodes set of a subtree. Choosing the unitary signals that represent each group as the roots of these subtrees generates a tree-structured constellation. The proposed tree search decoder decides to which sub tree the receive signal belongs by searching in the set of subtree roots. The final decision is made after a local search in the leaf nodes set of the selected sub tree. The adjacent subtree joint decoder performs joint search in the selected sub tree and its ＂surrounding＂ subtrees, which improves the Bit Error Rate （BER） performance of purely tree search method. The exhaustively search in the whole constellation is avoided in our proposed decoding algorithms, a lower complexity is obtained compared to that of Maximum Likelihood （ML） decoding. Simulation results have also been provided to demonstrate the feasibility of these new methods.
基金
Supported by the National Natural Science Foundation of China (No.60572148).
