Approximate ML Serial Decision Feedback Equalizer and Tomlinson-Harashima Pre-Equalization

Lingyang Song,Manav R. Bhatnagar,Qihao Li
2008-01-01
Abstract:In this paper, we study joint transmitter and receiver design problem for frequency-selective, time-invariant fading channels. We firstly propose a simple approximate maximum likelihood serial decision feedback equalizer (A-ML-SDFE) through decision feedback equalization, a Gaussian ap- proximation, a pre-whitening filter, and a matched filter. Secondly, assuming full channel knowledge available at the transmitter side, we perform pre-equalization in a downlink scenario by moving the decision feedback part of the A-ML-SDFE to the transmitter. The proposed A-ML-SDFE achieves much better performance than linear MMSE and MMSE-DFE with a lower complexity. The pre- equalizer further improves the system performance at low signal-to-noise ratio (SNR) with a reduced receiver complexity.
What problem does this paper attempt to address?