Differentiable Machine Learning-Based Modeling for Directly-Modulated Lasers

Sergio Hernandez Fernandez*, Ognjen Jovanovic, Christophe Peucheret, Francesco Da Ros, Darko Zibar

*Corresponding author for this work

Research output: Contribution to journalJournal articleResearchpeer-review

9 Downloads (Pure)


End-to-end learning has become a popular method for joint transmitter and receiver optimization in optical communication systems. Such approach may require a differentiable channel model, thus hindering the optimization of links based on directly modulated lasers (DMLs). This is due to the DML behavior in the large-signal regime, for which no analytical solution is available. In this letter, this problem is addressed by developing and comparing differentiable machine learning-based surrogate models. The models are quantitatively assessed in terms of root mean square error and training/testing time. Once the models are trained, the surrogates are then tested in a numerical equalization setup, resembling a practical end-to-end scenario. Based on the numerical investigation conducted, the convolutional attention transformer is shown to outperform the other models considered.

Original languageEnglish
JournalIEEE Photonics Technology Letters
Issue number4
Pages (from-to)266-269
Publication statusPublished - 15 Feb 2024


  • Directly modulated laser
  • Machine learning
  • Modeling
  • Optical communication
  • Transformer


Dive into the research topics of 'Differentiable Machine Learning-Based Modeling for Directly-Modulated Lasers'. Together they form a unique fingerprint.

Cite this