Machine Learning for Power Profiles Prediction in Presence of Inter-channel Stimulated Raman Scattering
- Rosa Brusin, A. M.
- Zefreh, M. Ranjbar
- Poggiolini, P.
- Piciaccia, S.
- Forghieri, F.
- Carena, A.
Publication date
January 2021
DOI Publisher
Institute of Electrical and Electronics Engineers (IEEE) Language
EnglishAbstract
Two artificial neural network (ANN) models are presented to predict power profiles over C+L–band in presence of inter-channel stimulated Raman scattering (ISRS) and to support non-linear interference (NLI) modeling. High prediction accuracy is obtained with maximum errors ≤ 0.1 dB over thousands different partial loads
Add to ORKG