Decision-feedback semiblind channel equalization in space-time coded systems

This paper addresses the problem of equalizing Multiple-Input Multiple-Output (MIMO) channels in Space-Time Coded (STC) systems. We propose to apply the Maximum Likelihood (ML) principle in order to compute the coefficients of a multidimensional Decision-Feedback Equalizer (DFE) that generalizes the classical DFE structure to the MIMO case. The equalizer coefficients are numerically computed by means of the Space Alternating Generalized Expectation-maximization (SAGE) algorithm. This algorithm is known to present local convergence problems due to the existence of multiple maxima in the likelihood function. To avoid this limitation, we suggest to incorporate a small number of a priori known symbols into the transmitted data. Since the aim of these symbols is not to improve performance but just to avoid misconvergence, we term the resulting equalizer as semiblind.