Fewer Parameter´s Novel Structure of Decision Feedback Equalizer For Nonlinear Channel

Based on analysis of cascades a linear filter with a memoryless nonlinear processor filter, combing characteristic of single layer neural network. Extended decision feedback equalizer´s structure with characteristic of DFE, two types novel structure of decision feedback equalizers are proposed in this paper, and adaptive algorithm is deduced by the LMS. Computer simulations show that PNNDFE which is one of the novel structure equalizers can availably remove nonlinear´s distortions and ISI, and BER´s performance also gets obviously improvement no matter what linear channel or nonlinear channel