Adaptive nonlinear equalizer with reduced computational complexity

This paper presents an approach of adaptive nonlinear equalization having much less computational complexity than existing nonlinear equalizers. The proposed equalizer first estimates many local regions in the symbol decision space, and then performs a biased finite impulse response filtering in each estimated local region, The final estimate of the transmitted symbol is obtained by a weighted average of the local results, By using a self-organizing and localized linear tuning methods, the proposed nonlinear equalizer can achieve an almost optimal decision making capability like the existing nonlinear equalizers, Simulation results show that the proposed equalizer can be effectively applied to on-line adaptation environments due to its reduced computational complexity, making complexity similar to that of a linear equalizer.