Overview of research on fiber nonlinear equalization algorithm based on artificial intelligence

Abstract

The necessity and importance of nonlinear equalization algorithm in optical transmission systems were outlined.The principle of classical nonlinear equalization algorithm was described,and the shortcomings and limitations of classical algorithms were presented.Combined with the status in recent years,four kinds of nonlinear equalization algorithms based on artificial intelligence were introduced in detail.These algorithms include artificial neural networks,support vector machine,unsupervised clustering and deep neural network.All the nonlinear equalization algorithms mentioned were compared in terms of performance,complexity,instantaneity and application flexibility.Finally,the future development trend of nonlinear equalization algorithm based on artificial intelligence was analyzed.

Keywords

• nonlinear equalization;artificial neural network;support vector machine;clustering;deep neural network