Improved soft isi cancellation for turbo equalization using full soft output channel decoder~s information

This paper presents an optimized turbo equalizer based on Laot's turbo equalization scheme by means of soft intersymbol interference (ISI) cancellation. Due to Berrou's turbo-principle, common turbo equalization schemes use the extrinsic information as a-priori input for the equalizer. In contrast to this, we propose the use of the full soft output channel decoder's information for the turbo interference canceller. The performance of this improved turbo equalizer by means of soft interference cancellation is evaluated in the environment of two channels, which introduce severe ISI. Simulation results show that the proposed turbo equalizer achieves a performance gain of up to 4 dB compared to Laot's original approach. The performance gain is achieved without any increase of computational complexity. Moreover, we apply a modified extrinsic information transfer (EXIT) chart to the proposed turbo equalizer and illustrate the convergence advantages in comparison to Laot's approach.