Low complexity linear MMSE equalization, channel decoding and estimation for frequency selective fast fading channels

Abstract

The present work focuses on low complexity (LC) Linear Minimum Mean Square Error (LMMSE) schemes in conjunction with the existing paradigm of turbo equalization (TE). As a consequence, the LC LMMSE equalizer was extended to new scenarios with unknown frequency selective fast fading Rayleigh channels. In this paper, the generalized Valenti and Woerner estimator is coupled with a LMMSE and LC LMMSE equalizer and used to accurately estimate the channel. Extrinsic information transfer (EXIT) charts are computed for the resulting TE to predict the decoding and equalization convergence behaviour. In addition, a comparison between TE using LC LMMSE and LMMSE equalizer is performed using bit error rate (BER) performance simulations.