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

Journal of Network and Systems Management

Deep Reinforcement Learning-Driven Optimization of End-to-End Key Provision in QKD Systems

Yeong-Jun Seok, Ju-Bong Kim, Youn‐Hee Han, Hyun-Kyo Lim, Chankyun Lee, Wonhyuk Lee

Published Feb 6, 2025• 6 citations
Publisher
DOI
10.1007/s10922-025-09902-7

Related papers

Deep Reinforcement Learning-Based Joint Security Provisioning in CV-QKD Over Optical Networks

Shifeng Ding, Chun-Kit Chan

Globecom•Dec 8, 2025•0 citations

AI-Driven Optimization of Quantum Key Distribution Protocols: a Reinforcement Learning Framework for Adaptive Quantum Cryptography

Maksud Khamidov

Lecture notes in networks and systems•Aug 2, 2026•0 citations

Deep-learning-based continuous attacks on QKD protocols

Théo Lejeune, François Damanet

Quantum Information Processing•Mar 26, 2026•2 citations

Key-Size-Driven Wavelength Resource Sharing Scheme for QKD and the Time-Varying Data Services

Jianing Niu, Yongmei Sun, Xianglong Jia +1 more

Integrating quantum key distribution (QKD) into the optical network through wavelength division multiplex (WDM) is promising for establishing large-scale QKD networks. However, the existing wavelength assignment schemes for WDM-QKD network all require to reserve a dedicate waveband for quantum signals, which influences the provisioning of data services, especially when the resource is tight. In this article, considering the quantum key pool (QKP) technology in the current QKD network, we propose a flexible key-size-driven wavelength assignment (KSD-WA) scheme to recycle the wavelength fragment to transmit quantum signals. Additionally, considering the noise impacts and the limited reconfiguration operations, we propose a heuristic algorithm to optimize the fragment selection with the objective of maximizing the secure key rate in the dynamic scenario. We also design a deep reinforcement learning (DeepRL) based algorithm to automatically learn the optimization policy under different environments. The simulation results verify that comparing with the conventional fixed-band (FB) scheme, the KSD-WA scheme assisted by the proposed optimization algorithms can provide adequate keys for encryption without affecting the provisioning of coexisting data services, which improves the compatibility of QKD with the real-life optical networks.

Journal of Lightwave Technology•Feb 3, 2021•16 citations

Hybrid Post-Quantum and Machine-Learning-Based Authentication for Secure QKD

Taranjeet Kaur, Gaurav Bathla, Anurag Singh

International Conference on Innovation & Growth in Next-Gen Intelligent Technology & Engineering (IGNITE)•Jun 1, 2026•0 citations

Deep Reinforcement Learning Based Topology Independent Routing and Load-Balancing in QKD Networks using Graph Neural Networks

Tim Johann, Sebastian Kühl, Stephan M. Pachnicke

Due to the unavailability of quantum repeaters, end-to-end secret keys need to be relayed in a hop-by-hop manner using so called trusted nodes in long-distance QKD networks. Therefore, a sufficient number of secret keys is required between all QKD devices along the path until the destination node is reached. As key generation rates are typically limited to a few kbps depending on link lengths, constraint-based routing algorithms considering key availability levels are desirable. We show that centralized routing based on Deep Reinforcement Learning (DRL) combined with Graph Neural Networks (GNNs) is able to reduce blockings and enhance load-balancing compared to conventional shortest path algorithms. Additionally, because of the training on random topologies, the agent is able to act in topologies unseen during training.

International Conference on Transparent Optical Networks•Jul 6, 2025•0 citations
Integrating quantum key distribution (QKD) into the optical network through wavelength division multiplex (WDM) is promising for establishing large-scale QKD networks. However, the existing wavelength assignment schemes for WDM-QKD network all require to reserve a dedicate waveband for quantum signals, which influences the provisioning of data services, especially when the resource is tight. In this article, considering the quantum key pool (QKP) technology in the current QKD network, we propose a flexible key-size-driven wavelength assignment (KSD-WA) scheme to recycle the wavelength fragment to transmit quantum signals. Additionally, considering the noise impacts and the limited reconfiguration operations, we propose a heuristic algorithm to optimize the fragment selection with the objective of maximizing the secure key rate in the dynamic scenario. We also design a deep reinforcement learning (DeepRL) based algorithm to automatically learn the optimization policy under different environments. The simulation results verify that comparing with the conventional fixed-band (FB) scheme, the KSD-WA scheme assisted by the proposed optimization algorithms can provide adequate keys for encryption without affecting the provisioning of coexisting data services, which improves the compatibility of QKD with the real-life optical networks.
Due to the unavailability of quantum repeaters, end-to-end secret keys need to be relayed in a hop-by-hop manner using so called trusted nodes in long-distance QKD networks. Therefore, a sufficient number of secret keys is required between all QKD devices along the path until the destination node is reached. As key generation rates are typically limited to a few kbps depending on link lengths, constraint-based routing algorithms considering key availability levels are desirable. We show that centralized routing based on Deep Reinforcement Learning (DRL) combined with Graph Neural Networks (GNNs) is able to reduce blockings and enhance load-balancing compared to conventional shortest path algorithms. Additionally, because of the training on random topologies, the agent is able to act in topologies unseen during training.