Research On Randomness Quantization And Real-time Post-processing In Quantum Random Number Generator

Randomness is an important resource in information and communication technology applications such as computer simulation,statistics,games and cryptography.For most of these applications,the quality of the random numbers is critical.For example,in modern cryptography,a poor quality of the random numbers used to generate encryption keys would lead to keys that could be predicted and it would pose a serious threat to the security of Internet communications.In recent years,with the development of quantum information technology,the inherent randomness of quantum mechanics makes quantum systems a perfect source of true random numbers.This paper focuses on the design and implementation scheme of a quantum random number generator based on vacuum fluctuation,which mainly includes the following aspects.(1)Implementation scheme of quantum random number generator based on vacuum fluctuation.We mainly study and implement a quantum random number generation device based on vacuum fluctuations.The device achieves randomness extraction from vacuum state quantum fluctuations,real-time random number post-processing using LFSR-based post-processing algorithms,and data storage and transmission functions for SD memory cards and gigabit networks,and achieves quantum random number generation at the level of 100 megabits per second,and passes the NIST randomness test.It is a design solution that can be used in practical quantum communication systems.(2)Non-uniform quantization scheme for randomness.In the process of implementing a quantum random number generation device based on vacuum fluctuation,we find that there is a contradiction between the high-speed sampling and the stable sampling in the traditional uniform quantization scheme using ADC or oscilloscope during the quantization of randomness.High sampling rate and high-resolution ADCs are costly and difficult to operate,while the upper limit of minimum entropy that can be extracted from each sampled data is found to be low by calculating the minimum entropy on the raw data.To solve this problem,we try to design a non-uniform quantization method using a multiplexed voltage comparator to realize the conversion from analog to digital signals through threshold comparisons,and compare it with the uniform quantization method under different QCNR,we find that the non-uniform quantization method has a certain improvement in the extractable minimum entropy under the case of larger QCNR.(3)Nonlinear feedback shift register(NLFSR)based post-processing algorithm for randomnumbers.For a linear shift feedback register-based post-processing algorithm,the linear complexity is equal to its order,while the feedback logic in it consists of binary addition(heterogeneous or operation),and the highest number of terms in its feedback expression does not grow.By introducing a local nonlinear feedback network,we design the NLFSR-based random number post-processing algorithm based on the original linear feedback shift register(LFSR)based algorithm.After simulation and verification,we find that the NLFSR-based post-processing algorithm can better remove the classical randomness from the original data and retain the quantum randomness while retaining the advantage that the LFSR algorithm is easy to implement in hardware.