Sang-Ho Shin and Kee-Young Yoo
Cellular automata, Pseudorandom number generator, Cryptography
Cellular automata (CA)-based pseudorandom number generators (PRNGs) have been proposed over the last 20 years. Although CA PRNGs in 2 or more dimensions provide the high randomness quality, the hardware complexity is inferior to 1-dimension CA PRNGs. However, the randomness quality of 1-dimension CA PRNG for non-linear CA rules is inferior to other CA PRNGs. In this paper, a novel CA PRNG based on concept of target & reference cells in 1-dimension is proposed. This concept is that the reference distinguish between target CA cells and reference CA cells. The complexity of the proposed scheme not only provides similar to 2-state, 3-neighborhood CA but reduce the correlation coefficient between the generated global states. To evaluate the randomness quality of the proposed method, the ENT and DIEHARD test suites are utilized. As the results, the randomness quality of the proposed scheme is superior to or equal to the previous works.
Important Links:
Go Back
