Periodic Reporting for period 1 - SEASON (SElf-mAnaged Sustainable high-capacity Optical Networks)
Berichtszeitraum: 2023-01-01 bis 2024-06-30
The targeted architecture considers joint MB and SDM (MBoSDM) and SDM-Passive Optical Networks (PON) networking, in terms of transmission and switching, while addressing innovation in terms of sliceable Bandwidth Variable Transceivers (S-BVTs) enabling Point-to-MultiPoint (P2MP. Along with the integration of (coherent) pluggable optical modules, smart Network Interface Cards (NICs) or the latest generation Data Processing Units (DPU) on open packet/optical white boxes, it will allow decoupling and providing open control and network management. Thus, such complex infrastructure requires rethinking the control and orchestration systems towards autonomous optical networks, adopting more agile DevOps methodologies. SEASON will leverage on cognitive networks powered by streaming telemetry, real-time network measurements and Artificial Intelligence/ Machine Learning (AI/ML)-aided service management for near-real time network operation, moving intelligence as close as possible to the data plane.
During the 1st period of the project modeling of the main data plane technologies was carried out namely for flexigrid, Multi-band and SDM as the most promising technology options to cost-efficiently scale network capacity and meet the increasing traffic demand. In addition, the first approach of the control plane was portrayed. This relies on NetDevOps approaches jointly with AI/ML in support of network operation and network orchestration, including Multi-Agent Systems (MAS). On top of this approach, AI/ML algorithms are applied for the near-real time control of network resources and services aiming at reducing energy consumption and ensuring perfosrmance, including moving intelligence as close as possible to the data plane, and devising a MAS distributed system,