Markdown Version | Session Recording

Session Date/Time: 02 Feb 2022 12:00

QIRG

Summary

This session featured a presentation by Mark Kaplan on Q-Line, a novel quantum network architecture developed at Very Cloud. The talk introduced Q-Line as a "quantum Ethernet" designed for local or metropolitan area networks, addressing scalability challenges and vulnerabilities associated with trusted nodes in traditional Quantum Key Distribution (QKD) networks. The discussion covered the technical architecture, security considerations, and various use cases, including integration into global quantum networks, secure data storage, inter-operator network connections, and future applications with quantum computers.

Key Discussion Points

• Introduction to Q-Line:
  • Presented by Mark Kaplan, Q-Line is a quantum network architecture aiming to provide long-term security based on quantum communication.
  • QIRG was reminded that it is a research organization, not a standards organization, with a focus on fostering collaboration and discussing research.
  • The core motivation is to address the scalability and security limitations of current QKD networks, particularly the expense of point-to-point QKD and the vulnerabilities introduced by trusted nodes for key routing.
• Q-Line Architecture and Protocol:
  • Q-Line decouples the operations of traditional QKD, introducing intermediate "Charlie" nodes (C1 and C2) between the source (Alice) and detector (Bob).
  • Alice generates a quantum state; C1 encodes information and chooses a basis; C2 chooses a measurement basis and re-injects randomness; Bob performs the final measurement. This randomness ensures privacy of the shared bit between C1 and C2.
  • The intermediate Charlie nodes are physically "lighter," utilizing standard photon modulators (telecom components) rather than costly single-photon detectors required at every node in traditional QKD.
  • The extremal nodes (Alice and Bob) are comparable to standard QKD devices (laser, modulators, single-photon detector).
  • Scalability Trade-offs: Q-Line is optimized for local/metropolitan areas. The total end-to-end distance (laser to detector) is limited, typically around 100 km, with 4-5 intermediate nodes being a practical regime. Each intermediate node (modulator) can induce additional loss.
• Q-Line vs. Traditional QKD Key Establishment:
  • In traditional QKD, adjacent devices establish keys in parallel, but routing keys between non-adjacent parties consumes primary keys and requires trusted nodes.
  • Q-Line uses time multiplexing (round-robin scheduling), meaning only one pair of nodes establishes a key at a time.
  • A key advantage of Q-Line is that it eliminates the need for key routing and trusted nodes for secure communication between any pair of connected nodes.
  • Under assumptions of uniform key distribution and detector cost dominance, the price per bit of key for Q-Line can be comparable to QKD.
• Security of Q-Line:
  • Theoretical Security: The security proof for Q-Line relies on established QKD security frameworks (e.g., by Toma, Mikhail, and Veri). The core argument is that an eavesdropper sees the same quantum state distribution in Q-Line as they would in a standard QKD protocol.
  • Side-Channel Attacks: Q-Line is compatible with decoy state protocols. Common side-channel attacks like Trojan horse attacks can be mitigated with standard techniques. More elaborate attacks, such as those relying on non-destructive photon counting, are an area of ongoing research and development.
• Use Cases and Applications:
  • Integration into Global Quantum Networks: Q-Line is envisioned as a "last kilometer" or "last mile" solution for distributing keys at the local area scale, integrating with larger backbones (e.g., in the EuroQCI project). It is compatible with other QKD implementations and can form building blocks for complex metropolitan architectures (e.g., deployments with Deutsche Telekom in Berlin and Orange in Paris).
  • Secure Data Storage: Q-Line can enhance secure storage solutions by combining QKD with classical cryptography and secret sharing schemes. By continuously re-encrypting shares distributed across multiple servers (potentially in cloud availability zones), it offers protection against data leakage and interception without relying on trusted nodes. It may also support quantum-assisted classical computation on encrypted shares.
  • Interconnecting Network Operators: Q-Line can be used to create overlapping network segments, providing multiple independent paths for key establishment between different operators without requiring a single trusted node at the interconnection point. This addresses trust issues between disparate operators.
  • Connection with Quantum Computers: A long-term vision involves connecting Q-Line networks to quantum computers to enable secure quantum cloud computing (verifiable and blind computation delegation). By replacing an end-node with a quantum computer, Q-Line could offer a scalable architecture for multiple clients to delegate quantum computations. Challenges include the need for true single-photon generators and frequency conversion.
• Discussion on Q&A:
  • Q-Line vs. Token Ring: The presenter clarified Q-Line can operate as a token ring but can also be combined with other techniques.
  • Q-Line vs. Optical Switches for QKD without Trusted Nodes: It was discussed that Q-Line maintains the advantage of adding nodes using standard telecom components. The advantage in terms of higher connectivity for Q-Line over optical switching for specific topologies would require further comparison.
  • Q-Line in Fully Quantum Networks (Entanglement): The presenter noted this is a complex question but highlighted that Q-Line implementations do not require quantum memories, simplifying complexity compared to entanglement-based networks.
  • Scalability Limits: The primary limit is optical fiber loss over the end-to-end distance, compounded by losses from intermediate modulators and fiber connections.
  • Adding Detectors to Intermediate Nodes: While potentially increasing robustness and security against certain attacks, adding detectors would make intermediate nodes more complex and costly, similar to full QKD devices or optical switching setups.
  • Core Advantage of Q-Line: The primary advantages are cost-effectiveness (using standard telecom components for intermediate nodes) and the ability to build a fully connected network without trusted nodes, which enhances security.

Next Steps

• Very Cloud is actively deploying and testing the Q-Line architecture in real-world field trials (e.g., with Deutsche Telekom and Orange).
• Developing and researching secure storage solutions leveraging Q-Line, with a goal for real-world implementation by early next year.
• Researching computation on stored data, focusing on classical and quantum-assisted classical computing.
• Continuing work towards the long-term vision of a quantum internet that supports secure quantum cloud computing algorithms.