A new blind equalization algorithm using higher order statistics in a decision feedback structure

The authors present a novel adaptive blind equalization algorithm named POTEA (power cepstrum and tricoherence equalization algorithm) using a nonlinear structure. It is based on the combined use of second- and fourth-order statistics of the received sequence. A nonlinear decision-feedback equalization (DFE) structure is proposed for equalizing nonminimum-phase channels with severe distortion, since a linear equalizer does not perform well in cases of channels with deep spectral nulls in their amplitude characteristics. The direct-form structure of the DFE consists of a feedforward T-spaced filter with an all-pass mixed phase transfer function, and a feedbackward T-spaced filter with a minimum phase one. The performance of POTEA for linear and decision-feedback equalizers is tested for the cases of 16-QAM (quadrature amplitude modulation) and 64-QAM.<<ETX>>