PRNG assessment tests based on neural networks
Artem A. Maksutov, Pavel N. Goryushkin, Alexey A. Gerasimov, Artem A. Orlov · 2018
The development of pseudo-random number generators (hereinafter called PRNG) is a very important direction in IT. This means that methods for assessing the efficiency of these generators' operating have a direct impact on their popularization and development in general. PRNG are used in many algorithms in software development, and the quality of these algorithms is responsible for the quality of the confidential data, and the quality of the software as a whole. However, existing assessing methods were created many years ago and do not take into account some of the properties that PRNG may have. This article is devoted to the development of new methods for PRNG analysis. The main objectives of this article are the implementation of the algorithm using artificial neural networks (hereinafter called as ANNs) and the study of the analyzer's properties.