Twilight – Twin Lightpaths

Twilight – Twin Lightpaths Project Overview

A digital-twin framework for full automation of disaggregated optical networks

The Twilight project develops an advanced digital-twin framework designed to fully automate the control and management of disaggregated optical networks. Leveraging contemporary machine learning and simulation techniques, Twilight integrates live telemetry data into predictive models, enabling network operators to simulate, validate, and optimize changes proactively. This ensures seamless operations, maximizes network utilization, and significantly reduces service provisioning times. Optical networks underpin more than 95% of global internet traffic, yet many still rely on manual processes and vendor-specific tools for network management, causing prolonged provisioning delays and suboptimal resource use. Twilight directly addresses these challenges by automating complex management tasks, reducing human error, and enabling real-time responses to network demands and disruptions, significantly enhancing network performance, reliability, and efficiency.

Project Team
Prof. Marco Ruffini
CONNECT Centre Profile

Prof. Dan Kilper
CONNECT Centre Profile

Dmitrii Briantcev, Research Fellow

Rana Kumar Jana, Research Fellow

Agastya Raj, PhD Student

Shuang Xie, PhD Student

Peyman Pahlevanzadeh, PhD Student

Collaborators

• HEAnet: Provides national network telemetry data and supports network trials.
• GÉANT: Facilitates pan-European backbone telemetry and supports cross-border network trials.
• BT: Engaged in operator use-case development and techno-economic analysis.

Funding Agency

Twilight is funded by Science Foundation Ireland under the Research Centres Programme and co-funded by the European Union Horizon Europe SNS JU

Location

Trinity College Dublin, Ireland