On the use of holonic agents in the design of information fusion systems

This paper describes an holonic interpretation of the JDL data fusion model and proposes a model of data and knowledge representation for an intelligent system in which the knowledge of the fusion agents and the imperfections of information are taken into account. Basic concepts borrowed from complex systems theory (holons, informons and holarchy) are described first then interpreted in a context of information fusion. The JDL model is being used to illustrate a holarchy of agents computing the various levels of processing of JDL. Finally, some examples of holons and informons are presented that are based on mathematical techniques identified as very important in any design of information fusion systems.