AI-driven digital twin for performance monitoring and fault prediction in optical communication networks

Abstract The rapid expansion of optical communication networks has increased the demand for intelligent monitoring, fault prediction, and adaptive network management techniques. Conventional optical network management approaches rely mainly on periodic measurements and predefined threshold-based decisions, which are unable to efficiently handle dynamic network conditions. This paper proposes an artificial intelligence (AI)-driven digital twin framework for real-time performance monitoring and fault prediction in optical communication networks. The proposed digital twin creates a virtual representation of the physical optical network by continuously collecting important network parameters, including received optical power, optical signal-to-noise ratio (OSNR), bit error rate (BER), quality factor (Q-factor), and transmission distance. An artificial neural network (ANN)-based prediction model is integrated with the digital twin to identify abnormal network conditions and predict possible performance degradation. The proposed approach is evaluated under different optical transmission scenarios by varying fiber length, noise level, and signal power. Simulation results demonstrate that the AI-enabled digital twin provides accurate prediction of optical link conditions and improves network monitoring capability compared with conventional machine learning approaches. The proposed framework achieves high prediction accuracy with low computational complexity, making it suitable for future intelligent optical communication networks. The proposed method achieves an accuracy of 98.9 %, demonstrating its effectiveness in monitoring network performance and accurately predicting faults in optical communication networks.