Research On Algorithms Of Randomness Detection

In computer science,random sequences have a wide range of applications,such as computer simulation,random sampling,numerical analysis,and computer programming.Especially in cryptography,there is a direct connection between randomness and the security of encryption.The most random sequences in practical applications are generated by pseudo-random number algorithms,which makes it necessary to use effective randomness detection algorithms to ensure the randomness of the generated sequences.The first Donald Knuth summarized 11 detection algorithms applied to real sequences,while nowadays 15 detection algorithms applied to binary sequences proposed by NIST(National Institute of Standards and Technology)are commonly used in cryptographic algorithms for cryptographic testing.In this paper,the first interval test algorithm is proposed by combining the interval test summarized by Donald Knuth.Since the specific steps of the original interval test need to ensure the existence of n intervals and can only be applied to real sequences,which leads to poor applicability,and the template-detection-related algorithm has the problem of time-consuming in practice.Therefore,the new algorithm solves the problem of low applicability of the original interval test and the efficiency of the template-detection-related algorithm,and applies it to binary sequences.Since encryption algorithms mostly generate 128 bit,192bit and 256 bit sequences,and the original algorithm cannot be applied to short binary sequences.In this paper,after further improvement of the first interval test algorithm,shorter sequences can be tested,and later the effectiveness and applicability of the new algorithm is proved by comparison experiments.After that,a random number generation and detection tool is designed under Windows 10 operating system using Py QT5 framework,which collects the true and false random number generation module,random number detection module,encryption algorithm random sequence generation module and short binary sequence detection module.Finally,five cryptographic algorithms,RC6,Rijindal,Twofish,SHA1 and SHA256,are tested using the NIST detection package with the improved first interval test algorithm to obtain the number of rounds of randomness preservation for these grouping or message digest cryptographic algorithms.The experimental results demonstrate that the coverage of NIST test package is increased by adding the first interval test algorithm without considering the sample error.