We are excited to share our recent contribution, “Anomaly Classification in CV-QKD over DWDM: Differentiating System Faults from Quantum Attacks”, presented at ICTON 2025 (July 2025, Barcelona).
In this paper we address the operational challenge of distinguishing between system faults (e.g., filter misalignment, component ageing) and malicious quantum attacks in continuous-variable quantum key distribution (CV-QKD) systems embedded in dense wavelength-division-multiplexing (DWDM) networks.
Key highlights:
- A framework that uses anomaly detection followed by classification to differentiate faults vs attacks in CV-QKD + classical channel coexistence.
- Evidence showing that traditional classifiers may not suffice, and that deeper ML (e.g., DNN) methods provide substantially higher accuracy.
- Practical relevance for SDN‐enabled optical networks planning to integrate quantum key distribution alongside high-capacity classical traffic.
Read the full paper here: https://doi.org/10.1109/ICTON67126.2025.11125358
We hope this work will serve as a valuable reference for network operators, quantum-secure communications researchers, and machine-learning practitioners in optical systems.