Mobile agents in adaptive hierarchical Bayesian networks for global awareness

In order to be efficient and robust, distributed computing applications must accommodate new data sources and alter their computational structure automatically. We demonstrate the utility of mobile agents for addressing such challenges in a distributed, real-time application. We describe our implementation of a distributed information fusion system based on Bayesian networks using the D'Agent mobile agent system. The Bayesian networks infer identity for clusters of vehicles using information from distributed sensors and databases.