Eavesdropping Attack And Defence On A Continuous-Variable Source-independent Quantum Random Number Generator

The rapid development of modern communication technology relies on efficient and secure communication encryption technology,which relies on high-speed and secure random number generation.Continuous variable Source-Independent Quantum Random Number Generator(CVSI-QRNG),It has obvious advantages in extracting true random numbers when the source is not trusted,and has made great progress in theoretical analysis and experimental implementation recently.By trusting the measurement device,even though the random source may be completely controlled by an eavesdropper,it can generate truly secure random numbers,which has great value.CV-SI-QRNG can theoretically generate true random numbers without making any assumptions about the source.Yet,in practical situation,the performance of the system will be affected by a mass of non-ideal factors with reducing the extractable randomness.Hence,the practical safety analysis of CV-SI-QRNG has attracted more and more attention.In addition,although there are many researches on improving the generation rate of CV-SI-QRNG random numbers,there are few innovations in algorithms.To address the two aforementioned practical issues,our main works of this paper are as follows:1.In view of the possible vulnerabilities in the actual CV-SI-QRNG system,a method of eavesdropping based on the fluctuation of the local intensity is proposed,and the performance comparison of the system before and after the attacking are analyzed.The results show that this attack scheme significantly reduces the randomness of the system.Based on this,we construct the defense means to resist such attacks,which provides reference for future practical experiments.2.Based on entropy uncertainty relation and extremum theorem,gauss state of continuous variable source has nothing to do with quantum random number generator safety analysis,theoretical simulation devices of various non-ideal factors effect on the system can extract the randomness,finally according to the performance of the various parameters in the model,select the optimal parameters,improve system performance,and using an optimized parameter set against defensive scheme;3.The permutation entropy change of quantum noise in CV-SIQRNG system was analyzed based on permutation entropy algorithm,and the simulation entropy rate changed with the embedding dimension,proving the possibility of permutation entropy algorithm applied to source-independent quantum random number generator.Efficient and secure random numbers have important applications in modern communication science technology and other fields.However,with the rapid development of quantum computers as well as mathematical algorithms,etc.,the security of the current algorithm-based generation of pseudo-random numbers is under serious threat.Quantum random number generators based on the principles of quantum physics are capable of generating unconditionally secure true random numbers.Among them,continuous-variable quantum random number generators have the advantages of high generation rate and low cost,which have become one of the current research hotspots.The theoretical security of the continuous variable quantum random number generator has been completely proved,however,there are still many security loopholes in the actual system.Researchers have proposed a scheme of source-independent quantum random number generator to address the security problem at the source side,which completely solves the security vulnerability of the source.Based on this paper,we further investigate the potential hacking problem of the rest of the continuousvariable source-independent quantum random number generator(CV-SIQRNG)system due to the non-ideal factors of the actual devices.The problem of eavesdropping against the fluctuation of the local oscillator intensity is proposed and solved,and the effect of quantum noise on the system is analyzed by applying the algorithm based permutation entropy to optimize the system parameters in order to improve the random number generation rate.To address the two aforementioned practical issues,our main works of this paper are as follows:1.Aiming at the possible vulnerability of the local oscillator in the actual CV-SI-QRNG system,an eavesdropping technique based on the fluctuation of the local oscillator intensity is proposed,and a practical security analysis based on the entropy uncertainty principle and extremality of Gaussian states is carried out,and theoretical simulations on the influence of non-ideal factors such as asymmetric beam splitter,finite sampling range and finite sampling resolution on the extractable randomness of the system are conducted.Using such eavesdropping attack,the extractable randomness of the system decreases significantly,which drops from 8.45bits to 0.It is proved that our proposed attack on the intensity fluctuation of the local oscillator severely damages the security of the CV-SI-QRNG system,resulting in the inability to generate secure random numbers.2.For the above proposed attack based on the intensity fluctuation of the local oscillator,the corresponding defense strategy is formulated in this paper in a targeted manner.According to the influence of non-ideal factors on the system in theoretical simulation and the change of extractable randomness before and after the system being attacked,the change of eavesdropper’s attack capability with non-ideal factors is analyzed,so as to select the best parameters to weaken the eavesdropper’s attack and improve the system performance.By setting the optimal sampling range,improving the sampling resolution,and adjusting the equipment to make it improve the transmission coefficient,the final extractable randomness is enhanced to guarantee the security performance of the actual experimental system and to maximize the resistance to hacker eavesdropping,so as to achieve the effect of defending against such local oscillator intensity attack and pave the way for future practical experiments.3.In this paper,the concept of permutation entropy is introduced in the CV-SI-QRNG system,and its calculation model is established in the system.On the basis of secure system,the paper analyzes the normalized permutation entropy with the change of quantum noise in the CV-SIQRNG system based on the permutation entropy algorithm,and theoretically simulates the entropy rate change of the average growth of permutation entropy with the embedding dimension.The simulation results demonstrate the possibility of applying the permutation entropy algorithm to CV-SI-QRNG to enhance the output rate of random numbers,which fills the gap in this field.