Improving Middle Square Method RNG Using Chaotic Map

Abstract

One of the classic approaches in PRNGs is the middle square method in which with a simple mathematical model generating pseudorandom numbers in high speed and minimum correlation between output numbers. Despite these unique characteristics, the method contains weaknesses that a broader application of this algorithm will face. In this paper is studied middle square method and then a logistic chaotic map is introduced with its specific features and its improved weaknesses via using these characteristics. Finally the NIST tests suite s are presented, in order to detect the specific characteristics expected from truly random sequences.