Compensation Technology Based On Digital Signal Processing In Conhenerent Optical Communication

Compared with the traditional optical communication system , coherent optical communca tion system with greater advantage, and with the development of the optical device made co he- rent optical communication system commercial applications for possible, so study com pensa- tion the dispersion of coherent optical communication system has the significant practical signifi cance. And with large-scale digital integrated circuit and application-specific integrated circuit, people pay more and more attention to the development of the optical signals using electrical signal transmission of equilibrium principle compensate, use digital signal processing method to compensate on the damage of can reduce cost dispersion, improve the signal quality, and has also has the ability of adaptive compensation.This paper mainly studies how to use digital signal processing technology to compensation dispersion and nonlinear in coherent optical communication system, to improve the performance of the system.First introduces the research background, After analysis the adaptive equalizer and adaptive algorithm. Analysis the structure of adaptive equalizer and classification them Compared the performance of minimum mean square algorithm LMS and least-square algorithm RLS. Through the simulation and analysis finished the performance comparision of RLS and LMS algorithm and found that RLS has better convergence and equilibrium effects. Secondly, the paper also introduces intelligent algorithm particle swarm optimization algorithm,apply PSO algorithm to optimize to optimize the equalizer halved halved, to compensate the fiber dispersion, Through simulation and analysis can get that PSO algorithm has a very good balanced performance last Through the compare convergence and simulation eye chart, ber of these three kinds of algorithm found that PSO algorithm has a better characteristics of LMS algorithm balanced effect than the RLS and LMS algorithm, although PSO algorithm has a bad convergence compare whit RLS algorithm, but, PSO algorithm is simple,and has less parameters, better convergence than LMS.