Reduced-complexity sequential particle belief propagation over correlated fading channels

In this paper, we design a factor graph based receiver for Bit-Interleaved Coded Modulation (BICM) over correlated fading channels. With the aim of enhancing the channel estimation accuracy, we propose belief propagation algorithm combined with particle filter as an alternative to typically used Kalman smoothing algorithm. To further reduce complexity, two schedules are proposed. One is employing mode and list of particle sequentially in iterations to update channel message. Another is an adaptive message update schedule in the demodulator and decoder. Simulation results illustrate performance improvement of the proposed algorithm and significant computational complexity reduction.