PQC-Enhanced QKD Networks: A Layered Approach

1. Problem Statement

The paper addresses the dual challenge of securing communications against future quantum attacks while overcoming the operational limitations of current Quantum Key Distribution (QKD) networks.

• The Threat: Current public-key cryptography (RSA, ECC) is vulnerable to Shor's algorithm on future quantum computers, necessitating a transition to Post-Quantum Cryptography (PQC).
• The Limitation of QKD: While QKD offers information-theoretic security based on physics, it is inherently limited to point-to-point links. Scaling to long distances requires "Trusted Nodes" (TNs), which act as intermediaries. These nodes introduce a trust assumption (they must be physically secured) and often require complex, separate Key Management Systems (KMS) and Software-Defined Networking (SDN) orchestration to forward keys, creating interoperability gaps and single points of failure.
• The Limitation of PQC: While PQC can be deployed end-to-end over existing infrastructure, its security relies on mathematical hardness assumptions that may eventually be broken. It also lacks the information-theoretic protection against "harvest-now, decrypt-later" attacks that QKD provides.
• The Gap: Existing hybrid solutions often require modifying network protocols, adding heavy KMS layers, or altering QKD hardware interfaces, hindering practical deployment.

2. Methodology: A Layered, KMS-Free Architecture

The authors propose a modular, three-layer architecture that unifies QKD and PQC without requiring a centralized Key Management System (KMS) for key forwarding or an SDN orchestrator. The design relies on compositional security where each layer operates independently.

The Three Layers:

1. Link Layer (Hop-wise Security):

   • Technology: WireGuard VPN tunnels.
   • Mechanism: Adjacent Trusted Nodes (TNs) establish point-to-point tunnels.
   • Key Source: Pre-Shared Keys (PSKs) for these tunnels are periodically rotated using symmetric key material retrieved from local QKD devices via the standardized ETSI GS QKD 014 interface.
   • Role: Secures the physical hop between nodes, ensuring confidentiality and authentication between directly connected TNs without needing a central KMS to forward keys.

2. Transport/Overlay Layer (End-to-End Security):

   • Technology: Rosenpass (an open-source PQC key exchange tool).
   • Mechanism: End nodes (e.g., Alice and Bob) perform a PQC key exchange over the chain of QKD-secured WireGuard tunnels.
   • Role: Establishes an end-to-end session key. Because the transport layer is already secured by QKD-derived keys, the PQC handshake is protected against passive eavesdropping.

3. Data Layer:

   • Technology: A final WireGuard tunnel for application data.
   • Mechanism: The session key generated by Rosenpass is injected as a PSK into this final WireGuard tunnel.
   • Role: Provides the actual secure channel for user data, achieving end-to-end confidentiality.

Key Design Features:

• KMS-Free: Removes the need for a network-level KMS to relay keys between hops. Local QKD KMSs handle key retrieval for immediate neighbors only.
• Fail-Safe Mechanisms: If QKD or PQC components fail, the system does not immediately drop connections. Instead, it injects random keys or terminates connections after specific grace periods (e.g., 60s–180s) to prevent downgrade attacks and ensure forward secrecy.
• Interoperability: Uses standard interfaces (ETSI-014) and unmodified network protocols (WireGuard), allowing integration with existing QKD hardware.

3. Key Contributions

1. Composable Layering: A novel framework combining QKD (for hop security) and PQC (for end-to-end security) without modifying QKD devices or network protocols.
2. Decoupled Forward Secrecy: Achieves forward secrecy at two levels:
   • Link Level: Periodic rotation of WireGuard PSKs via QKD.
   • End-to-End Level: Independent PQC key rotation via Rosenpass.
   • This limits the blast radius of a compromise; breaking one layer does not automatically compromise the other.
3. Implementation & Evaluation: Development of an open-source prototype using Arnika (a utility to inject QKD keys into WireGuard) and Rosenpass.
4. Operational Guidance: Defined migration paths for network operators to integrate these overlays into existing infrastructures.

4. Experimental Results

The authors validated the design through simulations and lab testbeds involving up to 100 trusted nodes.

• Scalability:
  • Setup time for a 100-hop network was 10.62 seconds, comparable to a 10-hop network (10.27s). This indicates that latency is dominated by the slowest single hop, not the cumulative number of hops.
  • The system successfully handled parallel PQC exchanges over multiple paths.
• Resource Efficiency:
  • Low control-plane traffic overhead.
  • A single Intel Core i9-13900H core successfully served 5,000 Rosenpass peers in simulation.
• Resilience:
  • Degraded Links: The system remained functional under high latency (300ms), jitter, and packet loss (1%), provided race conditions in the software were addressed.
  • Fail-Safe Behavior: When a QKD node was simulated to fail, the data tunnel remained active for approximately 548 seconds before gracefully degrading. This prevents immediate service disruption while ensuring that compromised keys are not used indefinitely.
• Security Validation:
  • The architecture successfully mitigated "harvest-now, decrypt-later" attacks by protecting the PQC handshake with QKD-derived keys.
  • An attacker would need to simultaneously break classical cryptography, PQC primitives, and compromise the physical Trusted Nodes to break the system.

5. Significance

This paper presents a pragmatic, deployable solution for Quantum-Safe Networks.

• Bridging the Gap: It solves the scalability issue of QKD (which usually stops at point-to-point) by using PQC for the end-to-end span, while solving the trust issue of PQC by using QKD to secure the transport path.
• No "Big Bang" Migration: Operators can deploy this overlay on existing QKD hardware and standard network equipment without replacing devices or rewriting core protocols.
• Defense-in-Depth: By layering QKD and PQC, the system creates a robust defense where the failure of one cryptographic assumption (e.g., a future break in PQC) does not lead to total system collapse, as the QKD layer still provides information-theoretic security for the hops.
• Standard Compliance: Adherence to ETSI GS QKD 014 ensures compatibility with the broader telecommunications ecosystem.

In conclusion, the authors demonstrate that a KMS-free, layered approach using WireGuard (secured by QKD) and Rosenpass (providing PQC) offers a scalable, low-latency, and highly secure path toward quantum-resistant communication infrastructure.