Adaptive nonlinear filtering techniques for data transmission

This paper describes the structure and the simulated performance of nonlinear adaptive filters appropriate for high speed serial data transmission over time-dispersive and slowly time-variant channels. The filters are designed to combat the two major obstacles in high speed digital signaling over such channels, namely, the presence of intersymbol interference due to the non-ideal frequency response of the channel and the slow, randomly time-variant response characteristics of the channel. Some results on the error rate performance of the adaptive filters, obtained by means of Monte Carlo computer simulation, are also presented.