— In this paper, we consider the joint transmitter and receiver design problem based on a novel approximate maximum likelihood decision feedback equalizer (A-ML-DFE) over frequency-selective, timeinvariant fading channels. By applying decision feedback equalization, Gaussian approximation, a pre-whitening filter, and a matched filter, the proposed A-ML-DFE can realize single symbol detection. The proposed scheme achieves near-optimal performance with a complexity lower than the linear MMSE and the MMSE-DFE. By assuming full channel knowledge, we perform pre-equalization and precoding in a downlink scenario at the transmitter side. The pre-equalizer can achieve perfect decision feedback and thus, provide better performance at low signalto-noise ratio (SNR) with a reduced receiver complexity. The precoder designed to minimize the analytical symbol error rate (SER) of this system can further improve the performance.
Lingyang Song, Are Hjørungnes, Manav R. Bha