Abstract

In nonlinear fiber optics, Modulation Instability (MI) is characterized by the emergence of new spectral components, either seeded from optical signals or appearing spontaneously from noise amplification. MI then constitutes an ideal playground for studying complex and incoherent nonlinear propagation dynamics [1]. In recent years, Artificial Neural Networks (ANN) have been used to characterize nonlinear dynamics and investigate ultrafast optical pulse evolution with remarkable accuracy and precision [2,3]. However, few works focused on incoherent optical propagation, with the identification of precursors and key-parameters responsible for the apparition of specific nonlinear phenomena in noise-driven systems [3-5].

© 2023 IEEE