Design of least-mean-square based adaptive optical equalizers

Abstract

Dispersion effects and attenuation over long distances cause degradation to a signal waveform in any communication channel, even optical fibers. Equalization is the process of restoring the shape of the signal waveform. Most current equalization techniques to combat signal distortion in optical fibers rely upon complex electronic realizations to process data converted from the optical signal of interest. The focus of this paper is to study the feasibility of designing efficient optical adaptive equalizers for the optical processing of guided lightwave signal waveforms corrupted by dispersion and attenuation effects varying slowly with time. The least-mean-square algorithm is used to adapt an equalizer's weights in real-time as the optical channel varies. The convergence and learning capabilities of the equalizer are analyzed as a function of the equalizer parameters and optical hardware errors. Optimal equalizer parameters are determined through analysis and numerical simulation such that the effects of optical errors and noise are reduced.