Deep Learning-Based Defect Identification for Quantum Key Distribution Devices

Quantum Key Distribution (QKD) technology, with its theoretically unconditional security, has demonstrated significant application value in secure communications system. However, under complex operating conditions characterized by strong electromagnetic interference and temperature/humidity changes in practical applications, QKD terminal devices are susceptible to defects such as quantum state preparation errors and decreased detector efficiency. These defects may cause key rate degradation, seriously threatening communication security. Therefore, accurate identification of QKD device defects is crucial. To address this issue, this paper proposes a deep learning (DL)-based defect identification framework for monitoring of QKD equipment operational status. The results demonstrate that the proposed deep learning algorithm exhibits remarkable advantages in complex system environments, achieving a defect identification accuracy of 99.7%. This work not only validates the effectiveness of deep learning algorithms in QKD device defect identification but also establishes a technical foundation for ensuring the stable operation of quantum-secured communication networks.