ANN-Based Mitigation of Optical Fiber Nonlinear Distortions in Data Center Networks

Abstract

This work presents a novel electrical equalization technique for mitigating the nonlinear distortions adhered to the received signals in data center networks. These distortions are due to the combination of the chromatic dispersion induced by the optical fiber channel and the direct detection system. An Artificial Neural Network (ANN) is trained to reverse the combined effect of the chromatic dispersion and direct detection.