A Novel Secure RNG Based On Three Entropy Sources

Abstract

Random numbers are essential for various cryptographic algorithms to make them work properly. Designing a random number generator (RNG) and estimating its entropy are crucial for evaluating its security. In this paper, we propose a new RNG based on three different entropy pools from the underlying computing machine. Entropy estimation based on a threat model is used to evaluate performance of the proposed RNG. Security and effectiveness of the proposed technique were substantiated with analysis’s results. Besides that, an exemplary developing scheme of pseudo-random number generators (PRNG) that were used from the proposed RNG are given.

Keywords: random number generators, entropy, min-entropy, threat-model, attacker, entropy source, Microsoft Threat Modelling Tool