Iterative receiver with Turbo equalization and soft demapping in multipath fading channels

In this paper, an iterative receiver with Turbo equalization and soft demapping is proposed for high data rate communication in multipath channels. Minimum mean square error (MMSE) equalization is adopted to avoid noise enhancement. With the assumption that the distribution of estimated symbols is Gaussian distribution, log likelihood ratios (LLR) of encoded bits are obtained with soft demapping. Using BCJR decoding algorithm, means and variances of transmitted symbols are calculated, and then fed back to the equalizer as priori information in the following iteration. The iterative receiver can achieve significant performance improvement than conventional ones without iteration, with a little increased computation complexity.