Mapping applications of cellular automata into applications of cellular automata networks

Cellular automata proved to be a promising model to simulate several complex systems: the requirement is that space and time, taken into account, have to be discretizable, while the system to be simulated has to satisfy locality and uniformity in the evolutionary space. Often, in dealing with the simulation of real complex systems some properties of locality are lost and consequently standard CA model application is very difficult. For this reason it is useful to extend the classical CA model and introduce feasible mechanisms in order to take advantage of the parallelism source of this computational model. With this aim the Cellular Automata Network (CAN) model was conceived that includes the advantages of classical CA models and introduces a new source of parallelism, i.e. the network of cellular automata. In this paper we deal with a sort of heuristics in order to map CA applications into CANs. This mapping can also be extremely useful as a proposal of a methodology to drive the modeling and simulation activity of complex phenomena that can be easily fragmented according to local interaction and components. (C) 2002 Elsevier Science B.V. All rights reserved.