Combining Machine Learning and the GN Model for Fast NLI Prediction in Dispersion-Managed Links

Non-linearity buildup in dispersion-managed network sections can be severe due to the small accumulated dispersion causing non-Gaussian signal distribution and correlation between noise contributions. We propose a machine learning-assisted, spatially disaggregated model able to provide a real-time and accurate non-linearity estimation in this scenario. ©2025 The Author(s)