Fiber Nonlinear Compensation Using Bi-directional Recurrent Neural Network Model Based on Attention Mechanism

In this paper, a bi-directional recurrent-neural network (Bi-RNN)- based algorithm with attention mechanism for nonlinear effects compensating in optical fibers is proposed for the first time. The numerical simulations show that the proposed scheme has 0.6 dB ~ 1.0 dB improvement compared with digital back propagation (DBP) algorithm for 1600 km transmission of 28Gbaud dual-polarization (DP) 16QAM signal.