Passive vs Active Approaches in Particle Approximations of Reaction-Diffusion Computing

Reaction-diffusion computing utilises the complex auto-catalytic and diffusive interactions underlying self-organising systems for practical computing tasks – developing variants of classical logical computing devices, or direct spatial embodiments of problem representations and solutions. We investigate the concept of passive and active approaches to reaction-diffusion computing. Passive approaches use front propagation as a carrier signal for information transport and computation. Active approaches can both sense and modify the propagation of the underlying carrier signal. Using particle approximations of reaction-diffusion behaviours in chemical wavefront systems, and the plasmodium of Physarum polycephalum, we demonstrate the similarities and differences between the two concepts. We provide examples of how both methods can be used for complex spatially represented computational tasks. We show that the active approach results in second-order emergent behaviours, exhibiting complex quasi-physical properties such as apparent surface tension effects and network minimisation which may have utility in future physical implementations of reaction-diffusion computing devices. [Article copies are available for purchase from InfoSci-on-Demand.com]