Author :
Gruver, William A. ; Marik, V.M. ; McFarlane, D.
Abstract :
Our environment is becoming more complex every day. The systems to be identified, described, managed, controlled, and maintained are becoming more distributed. The knowledge gained, stored and deployed to achieve intelligent behavior is growing in its volume and degree of distribution every hour. Both centralized decision making and control of distributed systems have their limitations and are not effective for large scale systems. The only solution is to decompose systems into smaller, efficient autonomous units responsible for local decision making and control, so that these units explore knowledge that is stored and maintained locally, and communicate among themselves only when needed. Such units, called agents, share the visions and goals, communicate, and negotiate in order to coordinate their activities and to cooperate in the full sense of the word.