Collision-based computing in biopolymers and their automata models

In collision-based computing, quanta of information are represented by auto nomous mobile signals. The signals travel in a uniform architectureless med ium. They collide to each other. Assuming the presence or abscence of a sig nal represent truth or falsity values of logical variable, we can consider logical functions are calculated at the sites of signals' collision. Physic ally, the signals are localized compact disturbances of medium's characteri stics. In this paper, we extensively exploit results published in Refs. 1-7 . We consider three types of localizations: breathers in one-dimensional ar rays of DNA molecules, excitons and groups of antialigned dipoles in two-di mensional arrays of Scheibe aggregates and microtubules respectively. Sever al forms of logical gates are extracted from published results on numerical simulation of breathers and excitons. In cellular automata models, we stud y interactions of the localizations with each other. We show what kinds of logical gates can be realized in such interactions. Parallels between physi cal and discrete automata models are provided.