BEM-based soft-in soft-out detection of DUSTM employing cyclic group codes

In this paper, the expectation-maximization (EM) algorithm for maximum a posteriori (MAP) estimation is employed to derive an iterative soft-in soft-out (SISO) detection scheme for differential unitary space-time modulation (DUSTM) system employing cyclic group codes. The data sequence produced by conventional differential detection is differentially recoded and used to estimate the channel state information which acts as the initial value for the Bayesian EM (BEM) algorithm. In each BEM iteration soft decision and more refined channel estimation are successively performed. In the final iteration, The transmitted symbol decisions can be made according to the MAP decision strategy and is differentially decoded to produce the output of the data sequence. Simulation results show the good performance of the proposed scheme compared with the conventional differential unitary space-time demodulation (DUSTD) and hard decision based iterative scheme.