Simultaneous gain profile design and noise figure prediction for Raman amplifiers using machine learning

A machine learning framework predicting pump powers and noise figure profile for a target distributed Raman amplifier gain is experimentally demonstrated. We employ single-layer neural network to learn the mapping from profiles figures. The obtained results show highly-accurate designs predictions, with maximum error on average of ~0.3dB. This provides comprehensive characterization thus valuable tool performance next-generation optical communication systems, expected amplification.