The convergence of intelligent software agents and field area networks

This paper discusses how intelligent software agents can be implemented on the nodes of a field area network (FAN). A number of aspects are to be mentioned to adapt this booming part of distributed artificial intelligence for FANs. First the typical properties of a FAN such as its small bandwidth and low computational power in the nodes restrict the possibilities for agent design. On the other hand, neural decision mechanisms and learning knowledge bases inside the agents or cooperation and negotiation among agents are a challenge for the FAN. This paper describes ways of combining both technologies and the benefits that emerge from this combination.