The CBEA algorithm: a cross-correlation approach to blind equalization and comparison with existing methods

A novel cross-correlation based framework is proposed for the problem of blind equalization. We assume that we have access to two observations, corresponding to the outputs of two channels excited by the same input. We propose a new algorithm which estimates the channels using as basic tool the phase of the cross spectrum of functions of the observations. The proposed method is computationally attractive, requires small input sample sizes, and performs well in low signal-to-noise ratios. Simulations show that the CBEA (cross-correlation blind equalisation algorithm) method outperforms existing methods at low SNR.