Quantum key distribution (QKD) provides unconditional security but relies on repeaters to extend coverage, thereby introducing repeater trust risks—compromised repeaters may leak keys. Brakerski/Fan-Vercauteren scheme (BFV)-based QKD addresses this issue through key encryption and quantum attack resistance. However, Fast Fully Homomorphic Encryption over the Torus (TFHE) outperforms BFV in encryption/decryption speed for single-qubit homomorphic XOR operations, which is critical for the real-time requirements of QKD. We propose TFHE-based QKD (TQKD), a quantum key distribution protocol based on public-key TFHE. During key forwarding, it leverages the “usable-but-unobservable” property of homomorphic encryption to prevent key exposure. A reduction proof verifies the scheme’s Indistinguishability under Chosen-Plaintext Attack (IND-CPA) security. To validate TQKD’s computational speed advantage, we developed code using the Open-Source Fully Homomorphic Encryption Library (OpenFHE) platform and designed single-pass and multi-hop relay experiments. We compared the computational efficiency of TQKD against QKD schemes based on similar homomorphic encryption algorithms, Brakerski-Gentry-Vaikuntanathan scheme (BGV) and BFV. Results demonstrate that our scheme achieves faster key encryption/decryption speeds in both scenarios, significantly reducing processing time compared to similar algorithms. Furthermore, while enhancing the scalability of quantum key distribution networks, the added computational overhead is negligible, indicating higher practical value.

Quantum Key Distribution (QKD) ensures secure key establishment through the principles of quantum mechanics; however, its effectiveness in practice hinges on dependable identity verification via classical channels during the post-processing phase. Current QKD implementations typically depend on pre-existing symmetric-key authentication, which suffers from limited scalability and complicated key management in extensive networks. Authentication methods utilizing post-quantum cryptography (PQC) signatures, based on complex mathematical assumptions, introduce extra and uncertain security dependencies, potentially compromising the security model integrity that QKD aims to maintain. This paper explores the application of hash-based signatures (HBS) for identity verification in the post-processing of QKD. HBS methods derive their security from cryptographic hash functions, which are integral to QKD protocols, allowing for scalable public-key-style authentication without the need for new computational assumptions. A detailed authentication framework is proposed, incorporating HBS-based verification into all essential phases of QKD post-processing, such as mutual certificate validation, basis sifting, parameter estimation, error correction verification, and privacy amplification. Security assessments indicate that the suggested framework maintains the security model integrity of QKD by relying cryptographically solely on the collision resistance of hash functions—without introducing new computational assumptions. At the system deployment level, it adheres to standard PKI trust assumptions which are necessary for public-key-style authentication and consistent with practical QKD network operations. Additionally, system-level evaluations affirm the scalability and practical applicability of HBS-based authentication, while also addressing the operational trade-offs among various HBS approaches in realistic QKD deployment contexts.