Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Log inSign up
Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Blog
Job Board
Q3AS

© 2026 Aqora Quantum S.A.S.

TermsPrivacyLegal Notice

IEEE International Conference on Advanced Networks and Telecommunications Systems

Neural network analysis in detection imperfections of continuous-variable quantum key distribution

Venkat Abhignan, Mohit Mittal, Megha Shrivastava, Abhijit Mitra

Published Dec 15, 2024• 2 citations
Publisher

Abstract

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.

DOI
10.1109/ants63515.2024.10898923

Related papers

Attack detection scheme for continuous variable quantum key distribution based on a graph neural network

Di Jin, Wenqi Jiang, Ying Guo +2 more

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.

Journal of the Optical Society of America B•May 3, 2024•5 citations

Homodyne-detector-blinding attack in continuous-variable quantum key distribution

Hao Qin, Rupesh Kumar, Vadim Makarov +1 more

We propose an efficient strategy to attack a continuous-variable (CV) quantum key distribution (QKD) system, which we call homodyne detector blinding. This attack strategy takes advantage of a generic vulnerability of homodyne receivers: A bright light pulse sent on the signal port can lead to a saturation of the detector electronics. While detector saturation has already been proposed to attack CV QKD, the attack we study in this paper has the additional advantage of not requiring an eavesdropper to be phase locked with the homodyne receiver. We show that under certain conditions, an attacker can use a simple laser, incoherent with the homodyne receiver, to generate bright pulses and bias the excess noise to arbitrary small values, fully comprising CV QKD security. These results highlight the feasibility and the impact of the detector-blinding attack. We finally discuss how to design countermeasures in order to protect against this attack.

arXiv:1805.01620•Physical Review A•Jul 12, 2018•117 citations

Practical Imperfections Affecting the Performance of CV-QKD Based on Coherent Detection

Nuno A. Silva, Daniel Pereira, Nelson Jesus Muga +1 more

Continuous variable-quantum key distribution (CV-QKD) based on coherent detection has potential to be a cost-effective solution to deploy secure keys over optical networks. This is due to its compatibility with mature optical communication technologies, such as homodyne and heterodyne detection schemes. Nevertheless, realistic implementations of CV-QKD systems suffer from imperfections in the practical devices used to generate and detect the coherent states. Those imperfections deteriorate the performance of the CV-QKD system and can be exploited by an eavesdropper to steal key information without being detected. In this work, we discuss the impact of practical device imperfections on the excess noise of the CV-QKD and on the achievable transmission distance. We also analyze the impact of detection imperfections on the measurement of the shot noise, and its impact on CV-QKD system security.

International Conference on Transparent Optical Networks•Jul 1, 2020•8 citations

Defending Against the Homodyne Detector-Blinding Attack on Continuous-Variable Quantum Key Distribution Using an Adjustable Optical Attenuator

Yijun Wang, Yanyan Li, Wenqi Jiang +1 more

A homodyne detector, which is also a common element in current telecommunication, is a core component of continuous-variable quantum key distribution (CV-QKD) since it is considered the simplest setup for the distinguishing of coherent states with minimum error. However, the theoretical security of CV-QKD is based on the assumption that the responses of the homodyne detector are always linear with respect to the input, which is impossible in practice. In the real world, a homodyne detector has a finite linear domain, so the linearity assumption is broken when the input is too large. Regarding this security vulnerability, the eavesdropper Eve can perform the so-called homodyne detector-blinding attack by saturating the homodyne detector and then stealing key information without being detected by the legitimate users. In this paper, we propose a countermeasure for the homodyne detector-blinding attack by using an adjustable optical attenuator with a feedback structure. Specifically, we estimate the suitable attenuation value in the data processing of CV-QKD and feed it back to the adjustable optical attenuator before the detector in real time. Numerical simulation shows that the proposed countermeasure can effectively defend against homodyne detector-blinding attacks and ensure the security of the Gaussian-modulated coherent state protocol with finite-size effect.

Entropy•Jun 13, 2025•0 citations

Saturation Attacks on DM CV-QKD: Threat Analysis and Machine Learning-Based Defense

Ye Li

Discretely modulated continuous-variable quantum key distribution (DM CV-QKD) protocols, while promising for extended transmission distances, face significant security threats in practice due to discrepancies between theoretical and actual system models. A critical vulnerability lies in the differential detector, which can be exploited by an eavesdropper to launch saturation attacks. In this paper, we conduct a comprehensive analysis of saturation attacks specifically targeting DM CV-QKD systems and propose a novel machine learning-based defense scheme utilizing the K-Nearest Neighbors (KNN) algorithm for attack detection. The simulation results confirmed that saturation attacks can undermine the actual security of the DM CVQKD system. The defense scheme proposed in this paper can accurately identify the probability of saturation attacks at 96.5%, and has been proven to be effective in defending against saturation attacks.

IEEE Access•Jan 1, 2026•0 citations

Quantum hacking: Saturation attack on practical continuous-variable quantum key distribution

Hao Qin, Rupesh Kumar, Romain Alléaume

We identify and study a security loophole in continuous-variable quantum key distribution (CVQKD) implementations, related to the imperfect linearity of the homodyne detector. By exploiting this loophole, we propose an active side-channel attack on the Gaussian-modulated coherent-state CVQKD protocol combining an intercept-resend attack with an induced saturation of the homodyne detection on the receiver side (Bob). We show that an attacker can bias the excess noise estimation by displacing the quadratures of the coherent states received by Bob. We propose a saturation model that matches experimental measurements on the homodyne detection and use this model to study the impact of the saturation attack on parameter estimation in CVQKD. We demonstrate that this attack can bias the excess noise estimation beyond the null key threshold for any system parameter, thus leading to a full security break. If we consider an additional criterion imposing that the channel transmission estimation should not be affected by the attack, then the saturation attack can only be launched if the attenuation on the quantum channel is sufficient, corresponding to attenuations larger than approximately 6 dB. We moreover discuss the possible countermeasures against the saturation attack and propose a countermeasure based on Gaussian postselection that can be implemented by classical postprocessing and may allow one to distill the secret key when the raw measurement data are partly saturated.

arXiv:1511.01007•Physical Review A•Jul 15, 2016•150 citations
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%90 \%90% in identifying the attack scenario utilizing a MATLAB neural network model.
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.
We propose an efficient strategy to attack a continuous-variable (CV) quantum key distribution (QKD) system, which we call homodyne detector blinding. This attack strategy takes advantage of a generic vulnerability of homodyne receivers: A bright light pulse sent on the signal port can lead to a saturation of the detector electronics. While detector saturation has already been proposed to attack CV QKD, the attack we study in this paper has the additional advantage of not requiring an eavesdropper to be phase locked with the homodyne receiver. We show that under certain conditions, an attacker can use a simple laser, incoherent with the homodyne receiver, to generate bright pulses and bias the excess noise to arbitrary small values, fully comprising CV QKD security. These results highlight the feasibility and the impact of the detector-blinding attack. We finally discuss how to design countermeasures in order to protect against this attack.
Continuous variable-quantum key distribution (CV-QKD) based on coherent detection has potential to be a cost-effective solution to deploy secure keys over optical networks. This is due to its compatibility with mature optical communication technologies, such as homodyne and heterodyne detection schemes. Nevertheless, realistic implementations of CV-QKD systems suffer from imperfections in the practical devices used to generate and detect the coherent states. Those imperfections deteriorate the performance of the CV-QKD system and can be exploited by an eavesdropper to steal key information without being detected. In this work, we discuss the impact of practical device imperfections on the excess noise of the CV-QKD and on the achievable transmission distance. We also analyze the impact of detection imperfections on the measurement of the shot noise, and its impact on CV-QKD system security.
A homodyne detector, which is also a common element in current telecommunication, is a core component of continuous-variable quantum key distribution (CV-QKD) since it is considered the simplest setup for the distinguishing of coherent states with minimum error. However, the theoretical security of CV-QKD is based on the assumption that the responses of the homodyne detector are always linear with respect to the input, which is impossible in practice. In the real world, a homodyne detector has a finite linear domain, so the linearity assumption is broken when the input is too large. Regarding this security vulnerability, the eavesdropper Eve can perform the so-called homodyne detector-blinding attack by saturating the homodyne detector and then stealing key information without being detected by the legitimate users. In this paper, we propose a countermeasure for the homodyne detector-blinding attack by using an adjustable optical attenuator with a feedback structure. Specifically, we estimate the suitable attenuation value in the data processing of CV-QKD and feed it back to the adjustable optical attenuator before the detector in real time. Numerical simulation shows that the proposed countermeasure can effectively defend against homodyne detector-blinding attacks and ensure the security of the Gaussian-modulated coherent state protocol with finite-size effect.
Discretely modulated continuous-variable quantum key distribution (DM CV-QKD) protocols, while promising for extended transmission distances, face significant security threats in practice due to discrepancies between theoretical and actual system models. A critical vulnerability lies in the differential detector, which can be exploited by an eavesdropper to launch saturation attacks. In this paper, we conduct a comprehensive analysis of saturation attacks specifically targeting DM CV-QKD systems and propose a novel machine learning-based defense scheme utilizing the K-Nearest Neighbors (KNN) algorithm for attack detection. The simulation results confirmed that saturation attacks can undermine the actual security of the DM CVQKD system. The defense scheme proposed in this paper can accurately identify the probability of saturation attacks at 96.5%, and has been proven to be effective in defending against saturation attacks.
We identify and study a security loophole in continuous-variable quantum key distribution (CVQKD) implementations, related to the imperfect linearity of the homodyne detector. By exploiting this loophole, we propose an active side-channel attack on the Gaussian-modulated coherent-state CVQKD protocol combining an intercept-resend attack with an induced saturation of the homodyne detection on the receiver side (Bob). We show that an attacker can bias the excess noise estimation by displacing the quadratures of the coherent states received by Bob. We propose a saturation model that matches experimental measurements on the homodyne detection and use this model to study the impact of the saturation attack on parameter estimation in CVQKD. We demonstrate that this attack can bias the excess noise estimation beyond the null key threshold for any system parameter, thus leading to a full security break. If we consider an additional criterion imposing that the channel transmission estimation should not be affected by the attack, then the saturation attack can only be launched if the attenuation on the quantum channel is sufficient, corresponding to attenuations larger than approximately 6 dB. We moreover discuss the possible countermeasures against the saturation attack and propose a countermeasure based on Gaussian postselection that can be implemented by classical postprocessing and may allow one to distill the secret key when the raw measurement data are partly saturated.