Multi-user decision-feedback space-time equalization and diversity reception

This paper investigates optimum infinite-length multi-user decision-feedback space-time equalization implemented with cross-feedback filters in order to remove not only post-cursor ISI but also post-cursor CCI. Closed-form expressions are derived for the optimal feedforward and feedback filters as well as the minimum achievable MSE (mean-square error) as a function of channel spectra for a one-antenna system in the ideal case, i.e. where cross-feedback filters are available for all interferers. Those results are then generalized with the help of a single-channel equivalent model to the multiple-antenna (space-time) case. Monte Carlo simulation results are provided comparing the MMSE obtainable with multi-user DFE, with standard DFE, with linear equalization (LE) and the matched filter bound in Rayleigh fading channels.