INFORMATION TECHNOLOGY FOR TESTING STATISTICAL REGULARITIES OF COMPUTER SOFTWARE GENERATOR

Abstract.In the paper, an overview of software generators of pseudorandom variables is carried out, the scope of application of pseudorandom variables is given. The total number of criteria for checking the statistical regularities of computer program generators was determined, and five criteria, which are most often used for checking in computational experiments, were studied. The conducted research confirmed the need to check the statistical regularities of computer generators, because this is especially important when applying them in computational experiments and applying the obtained results in various fields of knowledge (cyber security, simulation modeling, technical information protection systems, modeling of military operations, measuring equipment and computer computer games). Among such criteria are defined in the article: the Sherman, Chang-Spiring, Moran, Fronzini, and Greenwood-Quesenberry-Miller criteria. In the course of the research, it was determined that the following generally accepted requirements must be observed when designing and implementing generators of pseudorandom sequences: simplicity of hardware or software implementation; maximum speed; maximum approximation of the sequence obtained at the output of the generator to the theoretical law of distribution (not necessarily uniform); the ability to control output parameters; the possibility of operation of generators of pseudorandom sequences in a wide range of frequencies; the ability to quickly reconfigure its work depending on the choice of initial parameters. Thus, the development of information technology will allow to automate the process of checking statistical regularities of computer software generators and will provide an opportunity to ensure the adequacy and correctness of the results of statistical analysis.