Synergistic Interplay of Large Language Model and Digital Twin for Autonomous Optical Networks: Field Demonstrations

Yuchen Song,Yao Zhang,Anni Zhou,Yan Shi,Shikui Shen,Xiongyan Tang,Jin Li,Min Zhang,Danshi Wang

2024-11-01

Abstract:The development of large language models (LLM) has revolutionized various fields and is anticipated to drive the advancement of autonomous systems. In the context of autonomous optical networks, creating a high-level cognitive agent in the control layer remains a challenge. However, LLM is primarily developed for natural language processing tasks, rendering them less effective in predicting the physical dynamics of optical communications. Moreover, optical networks demand rigorous stability, where direct deployment of strategies generated from LLM poses safety concerns. In this paper, a digital twin (DT)-enhanced LLM scheme is proposed to facilitate autonomous optical networks. By leveraging monitoring data and advanced models, the DT of optical networks can accurately characterize their physical dynamics, furnishing LLMs with dynamic-updated information for reliable decision-making. Prior to deployment, the generated strategies from LLM can be pre-verified in the DT platform, which also provides feedback to the LLM for further refinement of strategies. The synergistic interplay between DT and LLM for autonomous optical networks is demonstrated through three scenarios: performance optimization under dynamic loadings in an experimental C+L-band long-haul transmission link, protection switching for device upgrading in a field-deployed six-node mesh network, and performance recovery after fiber cuts in a field-deployed C+L-band transmission link.

Networking and Internet Architecture,Optics

What problem does this paper attempt to address?

The problem this paper attempts to address is:

In autonomous optical networks, creating a high-level cognitive agent (i.e., "AI Agent") to achieve fully automated management is a challenge. Although large language models (LLMs) perform well in natural language processing tasks, they are insufficient in predicting the physical dynamics of optical communications, and directly deploying LLM-generated strategies into actual optical networks poses safety risks. Therefore, this paper proposes a scheme that combines digital twins (DT) and LLMs. By utilizing monitoring data and advanced models, DT can accurately characterize the physical dynamics of optical networks and provide dynamically updated information to LLMs, enabling reliable decision-making. Additionally, the strategies generated by LLMs can be pre-validated on the DT platform to ensure their safety and effectiveness.

Specifically, the paper demonstrates this synergy through the following three scenarios:
1. Optimizing performance under dynamic load in an experimental C+L band long-distance transmission link.
2. Protection switching during equipment upgrades in a field-deployed six-node mesh network.
3. Performance recovery after fiber cuts in a field-deployed C+L band transmission link.

These scenarios cover the main operations in the lifecycle of optical networks, including performance optimization during signal load fluctuations, protection switching during equipment upgrades, and performance recovery after failures. In this way, the paper demonstrates how DT-enhanced LLMs can solve these problems without human intervention, significantly advancing the autonomous management of optical networks.