Data-Driven and Intelligent Nonlinearity Equalizer Aided Optical Fibre Communication Networks

Abstract

Artificial intelligence techniques are outstanding in various fields because of their powerful performance and can offer improvements to optical communication systems. In optical fibre communication networks, we propose a data-driven and intelligent nonlinearity equalizer. This article mainly introduces applications of the first-order perturbation theory, principal components analysis (PCA) algorithms, iterative pruning method and transfer reinforcement learning. We note that these algorithms can effectively mitigate the nonlinear damage in optical fibre communication systems and can be quickly applied to different communication systems in different communication networks. We strongly believe that the artificial intelligence technology can solve existing problems of optical fibre communication systems, ease their bottlenecks, and play a vital role in the stable and rapid development of communication systems in the future.