The Pseudorandom Number Generator Generation Method with Genetic Programming for Lightweight Devices

Abstract

In this research a genetic programming (GP) method is proposed for producing pseudorandom number generators (PRNG) for lightweight devices especially wireless identification and sensing platform (WISP) family devices. These PRNGs arc produced with genetic programming methods using bit-entropy and serial correlation coefficient calculation as the fitness function. Qualities of these PRNGs are tested with the NIST statistical test suite that is a comprehensive tool that evaluates the statistical quality of the output of given PRNG. In addition a set of PRNG is generated with this GP method and the results are examined statistically.