From Reactive Multi-Agent models to Cellular Automata - Illustration on a Diffusion-Limited Aggregation model

Antoine Spicher (1) , Nazim A. Fatès (1) , Olivier Simonin (1)

1 MAIA - Autonomous intelligent machine

This paper deals with the synchronous implementation of situated Multi-Agent Systems (MAS) in order to have no execution bias and to allow their programming on massively parallel computing devices. For this purpose we investigate the translation of discrete MAS into Cellular Automata (CA). Contrarily to the sequential scheduling generally used in MAS simulations, CA is a model for massively parallel computing where the updating of the components is synchronous. However, CA expressivity is limited and not adapted to build models where independent entities may move and act on neighbor cells. After illustrating these issues on a simple example, we propose a generic method to translate discrete MAS into CA, called transactional CA. Our approach consists in translating MAS specified with the influence-reaction model into a transactional CA.

Mots clés: reactive multi-agent systems; cellular automata; influence-reaction model; diffusion-limited aggregation

Domaines: Automates cellulaires et gaz sur réseau [nlin.CG]

Antoine Spicher, Nazim A. Fatès, Olivier Simonin. From Reactive Multi-Agent models to Cellular Automata - Illustration on a Diffusion-Limited Aggregation model. 1st International Conference on Agents and Artificial Intelligence, Jan 2009, Portugal. http://www.icaart.org/. ⟨hal-00319471⟩