An Improved Particle Smoother for Blind Equalization in Time-Varying MIMO Channels

An improved fixed-lag particle smoother in the mixture Kalman filter (MKF) framework is developed for the blind equalization in time-varying multiple-input multiple-output (MIMO) channels. The improved particle smoother utilizes the uniform proposal distribution (PD) and the optimal resampling to generate particles. Compared to the particle smoother with the posterior PD, the improved particle smoother can provide more accurate estimation results, while both smoothers have same computational complexity. Simulation results are provided to illustrate the performance of the methods.