A Comparative Statistical Analysis of Pseudorandom Bit Sequences

Many users applied built-in random generator for their cryptography applications which is simple and fast. However, the randomness of generated pseudorandom numbers (PRNs) is under questioned whether it can support the reliable security in secure communication. In this paper, we examined two kinds of pseudorandom bit sequence (PRBS); conventional PRBS and chaos-based PRBS. Linear congruential method, Marsagliapsilas Ziggurat algorithm and Mersenne Twister method are used to generate the former, and the latter is produced using Chebychev map, logistic map and sawtooth-like map. To verify the randomness of the PRBS, two standard statistical tests FIPS140-2 and SP800-22 are performed. The results have been compared and analysed between conventional PRBS and chaos-based PRBS on their reliability as encryption key.