We are proud to announce the publication of our paper “Machine Learning-Driven Attack and Fault Classification in CV-QKD Systems After Anomaly Detection”, presented at ONDM 2025. The work explores the use of machine-learning techniques for classifying both system faults and intentional attacks within continuous-variable quantum key distribution (CV-QKD) systems after initial anomaly detection has flagged potential issues.
Key take-aways:
- Traditional classifiers (e.g., Gaussian Naive Bayes) perform inadequately in differentiating between faults and attacks in CV-QKD environments.
- Deep neural network (DNN) approaches show promise for accurate categorization of anomalous events in quantum-secure optical networks.
- The methodology contributes to enhancing the resilience of CV-QKD systems and supports broader integration of quantum-secure links in classical optical infrastructure.
Read the full paper here: https://doi.org/10.23919/ONDM65745
We believe this research will be of interest to network security professionals, quantum communications researchers, and developers working on converged quantum-classical optical systems.