Fig. 9
From: Field and lab experimental demonstration of nonlinear impairment compensation using neural networks
The measured performance of the NN-NLC algorithm at 2800 km. a The comparison of Q-factor improvement between Rx/Tx-side NN-NLC and Rx-side PPD. b The comparison of Q-factor improvement between Tx-side NN-NLC and filtered-DBP versus the number of multiplications
