Digital Twin-Assisted Lightpath Provisioning and Nonlinear Mitigation in C+L+S Multiband Optical Networks.

Area of Science:

• Optical Communications
• Network Engineering
• Artificial Intelligence

Background:

• Multiband (MB) optical transmission aims to boost network capacity.
• Nonlinear impairments (NLI) affect optical channels differently, complicating routing and spectrum assignment (RSA).
• Accurate quality of transmission (QoT) estimation is crucial for provisioning in MB networks.

Purpose of the Study:

• To propose a digital twin-assisted approach for lightpath provisioning in MB optical networks.
• To ensure required QoT in MB networks by integrating digital twins into the RSA problem.
• To develop a scalable and accurate QoT estimation tool for MB scenarios.

Main Methods:

• Development of the OCATA (Optical Channel Assignment Tool using AI) time-domain digital twin.
• Utilizing a Deep Neural Network (DNN) to model optical signal propagation and estimate QoT.
• A methodology to limit modeled channels for scalability while maintaining accuracy.
• Implementing NLI noise mitigation at the receiver during provisioning.

Main Results:

• OCATA-MB demonstrates remarkable accuracy in QoT estimation for all channels.
• The proposed approach reduces the blocking ratio by over 50% compared to traditional methods without NLI mitigation.
• Implementing NLI mitigation provides an additional over 50% reduction in blocking ratio.
• The digital twin approach effectively mitigates NLI noise with low complexity.

Conclusions:

• The OCATA-MB digital twin provides a complete solution for QoT-based lightpath provisioning in MB optical networks.
• This approach significantly enhances network performance by reducing connection blocking.
• The digital twin facilitates efficient NLI noise mitigation, improving overall network capacity and reliability.