Research On Reference-frame-independent Quantum Key Distribution

Cryptography plays a key role in today's information society.However,the security of the most widely used public-key algorithm relies on computation complexity.With the development of quantum computing technology,the security of classical cryptography based on public-key algorithm suffers from great threats.The first quantum key distribution(QKD)protocol,which was proposed by Charles Bennett and Gilles Brassard in 1984 and widely known as BB84,provides a new way for information security.Based on the laws of quantum mechanics,QKD allows two legitimate parties Alice and Bob to share secret keys.Combining the secret keys generated from QKD and the one-time pad method together,we can realize information-theoretic secure communications.However,real-life QKD systems are composed of practical devices which have some discrepancies with the theoretic model,and these may be exploited by Eve.Especially,in the source side of practical QKD systems,there are unavoidable state preparation flaws when preparing quantum states,intensity fluctuations and potential side-channel information when actively modulating different decoy states,and so on.The evasdropper can take advantage of these loopholes to launch corresponding attacks and compromise the practical security of QKD systems.At the same time,whether the performance of practical QKD systems can be further improved plays a vital factor for the large-scale application.Based on these questions,this dissertation focus on decoy-state BB84 QKD,reference-frame-independent QKD(RFI-QKD),and reference-frame-independent measurement-device-independent QKD(RFI-MDIQKD)to improve the practical security and performance of QKD systems,which are arranged as follows:1.Taking intensity errors and statistical fluctuations into consideration,we investigate the performance of biased decoy state BB84 QKD with parameter optimization.Simulation results show that parameter optimization can significantly improve not only the secure transmission distance but also the final key generation rate.Moreover,we realize an experimental decoy-state BB84 QKD demonstration with uncharacterized encoding sources and projective measurements,which improves the practical security and reduces the experimental complexity.By incorporating the mismatchedbasis data and considering the statistical fluctuation effects,we can distribute secret keys when the transmission distances of the standard fiber link are 101 km and 202 km,which demonstrates that a general BB84 QKD setting can distribute secret keys over long distances with relatively high practical security.2.We propose a biased decoy-state RFI-QKD scheme,where the basis choices of two legitimate parties are optimized to increase the fraction of matched-basis data.Simulation results demonstrate that this scheme can significantly improve the performance of RFI-QKD systems.Furthermore,to overcome the intensity fluctuation and the potential side-channel information of active intensity modulation in the general decoy-state RFI-QKD systems,we propose an efficient passive decoy-state RFI-QKD based on the parametric down-conversion source,where a beam splitter splits the idler pulses into four kinds of local detection events to estimate the parameters of single photon more accurately.Simulation results show that our scheme can obtain quite good performance even with the worst relative rotation of reference frames.3.We propose a loss-tolerant RFI-MDI-QKD protocol which is robust against state preparation flaws.Compared with the original six-state RFI-MDI-QKD,our scheme only needs four states,and these states can tolerate some preparation flaws,which improves the practical security of RFI-MDIQKD and simplifies the experimental implementation.Simulation results demonstrate that state preparation flaws in Z basis have adverse effect on the performance of loss-tolerant RFI-MDI-QKD while the source flaws in X and Y bases have almost no effect,which provides useful technical references for researchers to design RFI-MDI-QKD systems.