TeraFlow project was present at the 3rd ONFIRE Symposium, organised by CTTC, which was held online on February 24th, 2021 from 9:00 to 13:00 CET.

In this event, nine invited speakers presented and discussed the advances in Machine Learning applications and disaggregated optical networks, including:

  • Machine learning applications for physical layer security management, SDM networks and QoT estimation.
  • Machine Learning in mobile, multi-domain, multi-vendor ecosystems.
  • Optical white boxes design and GNPy.
  • Resilient disaggregated optical networks.


The presentation "Optical white boxes design and programmability adopting GNPy" held by Victor López from Telefónica I+D focused on showcasing work and findings partially supported by the TeraFlow project.

Abstract: White boxes ecosystem enables to decouple the hardware and operating system that is installed on it. The Telecom Infra Project is the fora where there is a definition of white boxes for optical networks. Even though there are advances in many aspects, there is still a lack of reference implementations with standard interfaces for programmability. There are gaps in devices models (like OpenConfig and OpenROADM) models for enabling physical impairment estimation engines (like GNPhy) as part of optical planning SDN NBI applications. This talk will present different architectural flavours which include TAPI and OC models that propagate upwards relevant information for physical impairment evaluation.