LMS-type self-adaptive algorithms for predictive decision feedback equalizer

Based on the fact that a predictive decision feedback equalizer (DFE) is composed of a feedforward filter (FFF) of a linear equalizer (LE), and a feedback filter (FBF) of a predictor, an LMS-type self-adaptive equalization method for the predictive DFE is proposed. The coefficients of the FFF and the FBF are estimated by applying a blind linear equalization algorithm and a decision feedback prediction algorithm, respectively. As the prediction algorithm provides superior convergence, the DFE always converges to a suboptimal point on the condition that the front-end LE converges. Simulation results demonstrate the effectiveness of the proposed method