|
Pseudorandom number generators (PRNGs) are widely used in many fields, especially in cryptographic applications. Pseudorandom number sequences generated via a poor PRNG will lead to weak or guessable its keys, and leak the information which is prevented. Based on the Golomb's assumptions on idea pseudorandom sequences and FIPS 140-2 randomness test criteria, this paper first presents a new approach for improving the randomness of pseudorandom sequences. Second, using a generalized synchronization theorem, the Henon map, the logistic map and a tube map constructs a new 8-dimensional chaotic generalized synchronization system (8DCGSS). Then using the 8DCGSS designs a chaotic PRNG (CPRN). The keyspace of the CPRNG is larger than 21117. Finally, using the FIPS 140-2 randomness test and a generalized FIPS 140-2 randomness test measures the randomness of the keystreams generated via the CPRNG, the Matlab PRNG, the RC4 algorithm, and the m-sequence, and the improved keystreams in term of our approach. The results show that our approach are able to increase significantly the randomness of the keystreams generated by the four PRNGs. |
|
Keywords:Probability and statistics; Improving randomness; Golomb's assumptions; Pseudorandom sequences; Chaotic pseudorandom number generator; RC4 algorithm; m-Sequence; FIPS 140-2 test |
|