Turbo equalization using particle filtering with grouping

This paper presents an efficient particle filtering algorithm for Turbo equalization. It approximates the maximum a posteriori (MAP) equalizer performance with low computational complexity. By grouping the identical trajectories of particles, we improve the efficiency of conventional particle filters and manage to work with large number of trajectories without a significant increase in complexity. EXIT chart analysis and Monte Carlo simulation confirm the excellent performance of the proposed method for channels even with severe inter-symbol interference.