Turbo equalization for EDGE: improved soft-ISI cancellation and its competitors

This paper presents a low complexity turbo equalizer for the EDGE transmission system based on improved soft intersymbol interference (ISI) cancellation. In contrast to Laot's original approach, we propose the use of the full soft channel decoder's information instead of the extrinsic information as a-priori input for the turbo interference canceller. The performance of our approach is evaluated in the environment of the hilly terrain (HT) fading channel. Simulation results show that the proposed turbo equalizer achieves a performance gain of up to 2.2 dB compared to a noniterative reduced-state maximum a-posteriori (MAP) equalizer in SNR-regions, where Laot's original turbo approach does not even converge. Furthermore, approach nearly achieves the same performance like the turbo equalizer based on MMSE linear filtering with time varying coefficients, despite its much lower computational complexity.