Improved turbo equalization based on soft ISI cancellation

Some turbo equalizers with perfect performance, such as the maximum a posteriori probability (MAP) equalizer and the minimum mean squared error (MMSE) turbo equalizer, are difficult to apply to certain wireless communication systems with large delay spread, due to their prohibitive complexity. By contrast, turbo equalization based on soft intersymbol interference (ISI) cancellation is much simpler; however, its performance is very poor. To avoid the shortcomings of the above equalizers, a novel equalization scheme is presented in this paper. The proposed method lies on an extension of soft ISI cancellation paradigm in the sense that the matched filter output is used to estimate the transmitter symbol. While the original method employs single output in the estimation, the number of outputs in the proposed method can vary depending on the delay spread of the channel. The extrinsic information transfer (EXIT) chart analyses and the simulation results show that the performance of the presented equalizer is much better than that of the original turbo equalization based on soft ISI cancellation, and it is only slightly more complex. The new turbo equalization scheme can perform as well as MMSE turbo equalization, but is simpler.