Development of a hardware cryptosystem based on a random number generator with two types of entropy sources

Abstract

In modern software, crypto-algorithms are widely used for both data encryption tasks, and authentication and integrity checks. There are well-known and proven crypto-algorithms. Their cryptoresistance is either mathematically proven or based on the need to solve a mathematically complex problem (factorization, discrete logarithm, etc.). On the other hand, in the computer world, information constantly appears about errors or «holes» in a particular program (including one that uses crypto-algorithms) or that it was broken (cracked). This creates distrust both in specific programs and in the possibility to protect something in general by cryptographic methods not only from special services, but also from ordinary hackers. A promising direction of research in this field is the implementation of a hybrid random number generator with two types of entropy sources in cryptosystems. The method and means of implementing a hybrid random number generator with two types of entropy sources: external – based on Zener diode noise and internal – based on the uncertainty state of the transistor-transistor logic structure are presented. One of the options for the practical implementation of a random number generator is presented, where two sources are used as a source of entropy: an external source – Zener diode noise and an internal source – the undefined state of the transistor-transistor logic structure. The functional diagram of the proposed random number generator with two types of entropy sources is given. The MATLAB/Simulink model of the proposed random number generator is built, the results of the statistical analysis of the generated random sequences by the NIST SP 800-22 test package are given.