Blind Separation Algorithm of Single Channel MPSK Signals based on MCMC

In order to cope with the performance bottleneck and the time complexity of particle filters algorithm for the blind separation of two MPSK mixture signals obtained by a single channel receiver, a novel algorithm based on Markov Chain Monte Carlo (MCMC) method is established which is based on the over-sampling signal model. By using over-sampling of the received signal, more information of waveform is utilized, so noise suppression is more effective. In the new algorithm the Gibbs sampling method is used to estimate the posterior probabilities of the symbols modulated in MPSK, so that the optimized Bayesian estimation can be obtained approximately. Furthermore, the unknown parameters in the signal model are iteratively estimated by the least Squares (LS) method. Theoretical analysis and the simulation results show the good performance and low complexity of the new algorithm.