Optimum Cellular Automata Configurations for Encryption

Abstract

Elementary cellular automata (ECA) is using Boolean logic functions to generate populations of ones and zeros. Population takes place temporally; each row of bits are populated based on the data from the previous row. The rules followed for generation of the next row are the functions mentioned above. Some rules produce very linear and predictable sets of binary data whereas some of them are chaotic in nature, like rule 30, and can be used to produce pseudo-random populations. This state by state calculation is ideal for cryptography in that it relies on an initial state (password or key) and can be streamed or generated as more data is transferred; the patterns are also largely unknown and are not predictable or linear in some cases. Currently, research shows many promising configurations of population generation that can be used for securing digital data or communications. Unfortunately the chaotic rules tend to repeat themselves after a given amount of temporal generation. This cyclic behavior is non-linear and near impossible to detect without generating the ECA. This research is aimed at finding the relationships between row length and cycle length in an effort to make a given key more effective by lengthening or eliminating this repetitious behavior.