Recursive estimator for blind MIMO equalization via BSS and fractional sampling

This paper addresses the problem of blind equalization based on a multi-input multi-output (MIMO) model. Using a more general structure for MIMO channels, we present a method that combines fractional sampling and a blind source separation (BSS) algorithm to recover transmitted symbols in the presence of ISI (intersymbol interference). Oversampling allows for converting the FIR MIMO model to an instantaneous mixing model. The equalization may then be performed via BSS. We propose a recursive estimator stemming from the Kalman filter for this task. The method achieves equalization even in the case of a slowly time-varying channel and in the presence of additive noise. Simulation results are presented illustrating the good performance of the method.