Memory-Saving Implementation of High-Speed Privacy Amplification Algorithm for Continuous-Variable Quantum Key Distribution

Privacy amplification is an indispensable procedure to share secure secret key between legitimate parties in quantum key distribution (QKD) because a portion of bits are still leaked to eavesdropper Eve after the secret reconciliation. Here, we propose a novel privacy amplification scheme with an optimal hash function. In particular, by utilizing the linear feedback shift register (LFSR) technique, we optimize and simplify the construction of the Toeplitz matrix. In this way, the LFSR-based Toeplitz matrix is used to execute the privacy amplification algorithm. The hardware storage resources can then be drastically saved, and the time consuming will be only about half of that without using the LFSR-based Toeplitz matrix. This work breaks the bottleneck of privacy amplification in high-speed continuous-variable QKD (CVQKD), and thus contributes to the real-time performance of the whole CVQKD system.