Scrambled Linear Pseudorandom Number Generators

F 2 -linear pseudorandom number generators are very popular due to their high speed, the ease with which a sizable state space can be created, and provable theoretical properties. However, they suffer from linear artifacts that show as failures in linearity-related statistical tests such binary-rank linear-complexity test. In this article, we give two new contributions. First, introduce transformations have been handcrafted good properties at same time programmable efficiently on superscalar processors, or even directly hardware. Then, describe some scramblers , is, nonlinear functions applied array reduce delete artifacts, propose combinations of extremely fast quality. A novelty our approach is use ideas theory filtered linear-feedback shift registers prove scramblers, rather than relying purely heuristics. end, provide simple, few hundred bits memory, properties, pass strong tests.