Experimental Demonstration of Online Learning-Based Concept Drift Adaptation for Failure Detection in Optical Networks

2026-02-11

AI Summary

This work proposes an adaptive fault detection method based on online learning to address the performance degradation of static models in optical network fault detection caused by concept drift. By introducing an online learning mechanism into optical network fault diagnosis for the first time, the approach integrates concept drift detection with dynamic model updating to enable real-time adaptation to environmental changes. Experimental results demonstrate that, compared to conventional static models, the proposed method achieves up to a 70% improvement in fault detection performance while maintaining low latency, significantly enhancing both adaptability and robustness.

Technology Category

• Machine Learning: Online Learning & Bandits
• Search and Optimization: Learning to Search
• Computer Vision: Learning & Optimization for CV

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Systems and Infrastructure for Web, Mobile and WoT: Experiences and lessons learnt from Web-based algorithms and system deployments
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

We present a novel online learning-based approach for concept drift adaptation in optical network failure detection, achieving up to a 70% improvement in performance over conventional static models while maintaining low latency.

Problem

• concept drift
• failure detection
• optical networks
• online learning

Innovation

• online learning
• concept drift adaptation
• failure detection
• optical networks
• low latency