Fiber Nonlinearity Mitigation for mQAM Coherent Optical Communication System Utilizing Random Forest Algorithm

In this paper, the random forest approach is used to mitigate fiber nonlinear damage in a coherent optical communication system. The feasibility of the proposed method is verified in both uniformly distributed m-level quadrature amplitude modulation (mQAM) and probabilistic shaping (PS) mQAM long-haul coherent optical communication system. All of the simulation findings indicate that the tolerance of nonlinearity can be improved by applying this algorithm. As a result, the optimal incoming fiber power can be increased in contrast with using only chromatic dispersion compensation (CDC).