Open AccessTesting a hardware random number generator using NIST statistical test suite
Author(s) -
М. О. Пикуза,
С. Ю. Михневич
Publication year - 2021
Publication title -
doklady belorusskogo gosudarstvennogo universiteta informatiki i radioèlektroniki
Language(s) - English
Resource type - Journals
eISSN - 2708-0382
pISSN - 1729-7648
DOI - 10.35596/1729-7648-2021-19-4-37-42
Subject(s) - nist , random number generation , generator (circuit theory) , computer science , self shrinking generator , test suite , noise (video) , random seed , pseudorandom binary sequence , convolution random number generator , computer hardware , pseudorandom number generator , random sequence , shift register , binary number , algorithm , random function , electrical engineering , mathematics , engineering , statistics , test case , random variable , telecommunications , arithmetic , artificial intelligence , image (mathematics) , mathematical analysis , chip , wind power , power (physics) , regression analysis , quantum mechanics , machine learning , induction generator , physics , distribution (mathematics) , natural language processing
Random number generators are required for the operation of cryptographic information protection systems. For а correct application of the generator in the field of information security, it is necessary that its output sequence to be indistinguishable from a uniformly distributed random sequence. To verify this, it is necessary to test the generator output sequence using various statistical test suites such as Dihard and NIST. The purpose of this work is to test a prototype hardware random number generator. The generator is built on the basis of the ND103L noise diode and has a random digital sequence of binary numbers at the output. In the prototype there is a possibility of regulating the amount of reverse current through the noise diode, as well as setting the data acquisition period, i.e. data generation frequency. In the course of operation, a number of sequences of random numbers were removed from the generator at various values of the reverse current through the noise diode, the period of data acquisition and the ambient temperature. The resulting sequences were tested using the NIST statistical test suite. After analyzing the test results, it was concluded that the generator operates relatively stably in a certain range of initial parameters, while the deterioration in the quality of the generator's operation outside this range is associated with the technical characteristics of the noise diode. It was also concluded that the generator under study is applicable in certain applications and to improve the stability of its operation, it can be improved both in hardware and software. The results of this work can be useful to developers of hardware random number generators built according to a similar scheme.