2-D CA Variation With Asymmetric Neighborship for Pseudorandom Number Generation

Optimization of 2d Lattice Cellular Automata for Pseudorandom Number Generation

“…In previous works [12][13][14]19 where n is the total number of CA cells. This means that a string of n bits is produced by the CA X(t) at a particular discrete time step t. If the strings of n bits for all time steps are juxtaposed, thus forming t*n bits, a general equation of a CA can be defined as, …”

“…Recently, 2-d CA variation 19 with asymmetric neighborship properties was introduced. Asymmetric neighborship property means that a certain cell a considers cell b as its neighbor but cell b does not consider cell a as neighbor.…”

“…The structure of a 2-d CA variation 19 [16][17][18] . Initial experiments show that there are better 2-d CA structures other than these two, e.g., when 2<VL<L. This 2-d lattice CA property is further explored in the later sections of this paper.…”

“…Thus, it is considered necessary to design a CA PRNG with the least structural complexity and a cost-effective implementation. randomness performance [12][13][14]19 . This paper keeps the same objective in focus.…”

Comparison between 2D cellular automata based pseudorandom number generators

“…Under this observation, in this paper, we evaluate the nature of 4-neighbor neighborship as a mean to improve the randomness quality. A 2D CA configuration that keeps most of the parameters proposed in [2] was used, except that different assymetric 4-neighbor neighborhoods were explored [3]. The four neighbors were restricted to be located in a 5 × 5 area around a center cell in order to reduce the searching space.…”