A Machine Learning Method for Fiber Nonlinear Compensation

As the optical system capacity increases rapidly, fiber nonlinear effect compensation has been considered as a key technology. In this work, a new nonlinear effects compensation method based on machine learning is proposed. We first convert the data into the patterns to compensate for the nonlinear inter-symbol interference (ISI). As a nonlinear tool, the artificial neural network (ANN) is trained by preprocessed patterns with supervision. Then support vector machine (SVM) is used to work as a classifier.