Enhanced Chaos-based Pseudo Random Numbers Generator

Abstract

Enhancing chaos-based PRNGs quality is the subject of this paper. The big problem that chaotic systems suffer from is the limited arithmetic computation size, that is; dynamical degradation in digital chaotic systems appears due to the digitization process. Quantization error, limited cycle-length and worse statistical proprieties of the chaotic sequence are in fact the most shortcomings of implementing chaotic systems on digital computers. In this paper, we have proposed a simple method for purifying as maximum as possible digital chaotic systems from aforementioned shortcomings. The method consists of using a mechanism to perturb the chaotic orbits during randomly chosen periods, and in the same time using another circuit to convert the chaotic sequence to another one that provide good statistical proprieties. The proposed scheme has been applied on the Logistic chaotic map using low arithmetic computation size, the results in terms of cycle-length and uniformity showed that the proposed scheme is so better compared to the original system and can extremely enhance the chaotic proprieties of a given digital chaotic system.