Blind adaptive degenerated look-up table based perturbative nonlinear compensation for 16QAM probabilistically shaped signals

A degenerated look- up table based perturbative nonlinearity compensation (DLUT-PNC) algorithm for probabilistically shaped 16QAM signals was proposed to compensate intra-channel fiber nonlinearity. However, this method adopts the standard 16 QAM signals without considering the influence of ASE noise to calculate the degenerated elements of table. Hence, the degenerated elements of the table are not strictly optimal. In this paper, a blind adaptive DLUT-PNC (BADLUT-PNC) method based on the gradient descent algorithm is proposed. We use the gradient descent algorithm to optimize the degenerated elements and then obtain the optimal degeneration scheme. In a single channel 70GBaud dualpolarization 16QAM transmission simulation with a 1200km link, the proposed scheme is investigated. The simulation result shows that the extra 0.15 similar to 0.45dB SNR improvements can be achieved by adopting our proposed BA-DLUT-PNC compared to the conventional DLUT-PNC method.