Blind channel estimation for continuous‐variable quantum key distribution
Geng Chai, Dengwen Li, Zhengwen Cao, Minghui Zhang, Peng Huang, Guihua Zeng
Abstract
For the maximum-likelihood parameter estimation (MPE) of continuous-variable quantum key distribution (CVQKD), it sacrifices partial key when estimating the channel parameters, and when the channel is varying, even slowly, legitimate communication parties must constantly exchange partial input signal and output signal to update the channel estimation. Therefore, the MPE leads to the following problems in practical application: (a) MPE achieved by randomly sampling the states can not accurately reflect the dynamic channel; (b) there exist a trade-off between the secret key rate (SKR) and the accuracy of parameter estimation in the finite-size regime; and (c) information leakage may be caused by the correlation between key variables. In this article, a novel parameter estimation method for CVQKD is proposed without affecting the security, and the consistency between the method and the physical model is given. Experimental simulation results show that the method can effectively improve the SKR and secure propagation distance. The experimental verification of the proposed method is given through the fiber system. Besides, the security analysis demonstrates that the method is secure against entanglement-distillation attack.