Misdiagnosing faults as attacks significantly reduces the key-generation uptime of CV-QKD systems. To address this, we characterize attacks and faults in CV-QKD systems and propose an approach for anomaly detection and its classification as faults and attacks in CV-QKD copropagating within dense wavelength division multiplexing (DWDM) systems. The proposed method combines analytical techniques for anomaly detection with a deep neural network (DNN) for accurate classification. This approach minimizes unnecessary system shutdowns and can be seamlessly integrated into SDN-enabled CV-QKD systems. Results demonstrate that traditional classifiers, such as Gaussian Naive Bayes (GNB), are insufficient for this task. Instead, the use of a DNN, despite its fourfold increase in prediction time, ensures significantly higher classification accuracy, thereby enhancing system reliability and operational efficiency.
Misdiagnosing faults as attacks significantly reduces the key-generation uptime of CV-QKD systems. To address this, we characterize attacks and faults in CV-QKD systems and propose an approach for anomaly detection and its classification as faults and attacks in CV-QKD copropagating within dense wavelength division multiplexing (DWDM) systems. The proposed method combines analytical techniques for anomaly detection with a deep neural network (DNN) for accurate classification. This approach minimizes unnecessary system shutdowns and can be seamlessly integrated into SDN-enabled CV-QKD systems. Results demonstrate that traditional classifiers, such as Gaussian Naive Bayes (GNB), are insufficient for this task. Instead, the use of a DNN, despite its fourfold increase in prediction time, ensures significantly higher classification accuracy, thereby enhancing system reliability and operational efficiency.
As continuous-variable quantum key distribution (CV-QKD) systems move toward integration with dense wavelength division multiplexing (DWDM) networks, distinguishing between external faults and malicious attacks becomes crucial for maintaining secure and high-availability key generation. While most existing studies focus on either attack mitigation or fault resilience independently, little attention has been paid to their joint classification. In this work, we present a machine learning-based approach to differentiate between system faults and quantum-level attacks in CV-QKD systems operating under DWDM coexistence. We emulate two practical filter-related faults—shifted filter frequency (SFF) and narrow filter bandwidth (NFB)—and a power-splitting channel tampering attack. Using Bob’s quadrature data (pB, qB) collected from a CV-QKD simulation test setup, we evaluate five classifiers: Gaussian Naive Bayes (GNB), K-Nearest Neighbors (KNN), Support Vector Machine (SVM), Random Forest (RF), and Deep Neural Network (DNN). We assess performance across various input sample sizes, measuring both classification accuracy and average prediction time. Our results show that while GNB offers the fastest predictions, it suffers from low accuracy. In contrast, DNN achieves near-perfect classification accuracy but at higher computational cost. RF presents a favorable trade-off, offering high accuracy with low latency. This work enables intelligent diagnosis of faults and attacks in telecom-integrated QKD systems, improving the uptime without compromising security.
The practical security of a continuous-variable quantum key distribution (CV-QKD) system is vulnerable to various attack strategies due to the significant difference between the idealized theoretical model and the practical physical system. The existing countermeasures against these attacks involve exploiting different real-time monitoring modules, which presents a challenge in effectively classifying attacks. We investigate a graph neural network (GNN)-based attack detection scheme for CV-QKD, which models data as a graph structure using three different methods for various conditions. Particularly, one of the proposed methods requires no additional devices and can detect attacks with over 99% accuracy. The algorithm can be expanded to different scenarios without additional training and can achieve a detection efficiency of more than 95%. Furthermore, our proposed scheme incorporates anomaly detection algorithms into the detection module, enabling 85% effective detection of partially unknown attacks with minimal security data.
Despite significant advancements in continuous-variable quantum key distribution (CV-QKD), practical CV-QKD systems can be compromised by various attacks. Consequently, identifying new attack vectors and countermeasures for CV-QKD implementations is important for the continued robustness of CV-QKD. In particular, as CV-QKD relies on a public quantum channel, vulnerability to communication disruption persists from potential adversaries employing denial-of-service (DoS) attacks. Inspired by DoS attacks, this paper introduces a threat in CV-QKD called the channel amplification (CA) attack, wherein Eve manipulates the communication channel through amplification. We specifically model this attack in a CV-QKD optical fiber setup. To counter this threat, we propose a detection and mitigation strategy. Detection involves a machine learning (ML) model based on a decision tree classifier, classifying various channel tampering attacks, including CA and DoS attacks. For mitigation, Bob, postselects quadrature data by classifying the attack type and frequency. Our ML model exhibits high accuracy in distinguishing and categorizing these attacks. The CA attack's impact on the secret key rate (SKR) is explored concerning Eve's location and the relative intensity noise of the local oscillator (LO). The proposed mitigation strategy improves the attacked SKR for CA attacks and, in some cases, for hybrid CA-DoS attacks. Our study marks an application of both ML classification and postselection in this context. These findings are important for enhancing the robustness of CV-QKD systems against emerging threats on the channel. Published by the American Physical Society 2024
Deep neural networks (DNNs) have been employed in continuous-variable quantum key distribution (CV-QKD) systems as attacking detection portions of defense countermeasures. However, the vulnerability of DNNs leaves security loopholes for hacking attacks, for example, adversarial attacks. In this paper, we propose to implement the one-pixel attack in CV-QKD attack detection networks and accomplish the misclassification on a minimum perturbation. This approach is based on the differential evolution, which makes our attack algorithm fool multiple DNNs with the minimal inner information of target networks. The simulation and experimental results show that, in four different CV-QKD detection networks, 52.8%, 26.4%, 21.2%, and 23.8% of the input data can be perturbed to another class by modifying just one feature, the same as one pixel for an image. We carry out this success rate in the context of the original accuracy reaching up to nearly 99% on average. Further, by enlarging the number of perturbed features, the success rate can be raised to a satisfactory higher level of about 80%. According to our experimental results, most of the CV-QKD detection networks can be deceived by launching one-pixel attacks.
The security of the continuous-variable quantum key distribution (CVQKD) system is subject to various attacks by hackers. The traditional detection method of parameter estimation requires professionals to judge known attacks individually, so the general detection model emerges to improve the universality of detection; however, current universal detection methods only consider the independent existence of attacks but ignore the possible coexistence of multiple attacks in reality. Here, we propose two multi-attack neural network detection models to handle the coexistence of multiple attacks. The models adopt two methods in multi-label learning: binary relevance (BR) and label power (LP) to deal with the coexistence of multiple attacks and can identify attacks in real-time by autonomously learning the features of known attacks in a deep neural network. Further, we improve the model to detect unknown attacks simultaneously. The experimental results show that the proposed scheme can achieve high-precision detection for most known and unknown attacks without reducing the key rate and maximum transmission distance.
Hacking vulnerabilities in continuous-variable quantum key distribution (CV-QKD) systems are investigated, focusing on practical imperfections in homodyne detection. Specifically, these flaws can be exploited in detector-blinding and detector-saturation attacks, where an eavesdropper can undermine secure communication, especially in CV-QKD. Further, a neural network-based detection method is implemented as a potential countermeasure to classify homodyne detector output voltage data, aiding in identifying such potential hacking scenarios. We obtain an overall accuracy of over 90% in identifying the attack scenario utilizing a MATLAB neural network model.