Towards Smart Intelligent Agents in the Command and Control Environment
Abstract
Successful implementation of Software Agents (SAs) depends on modeling the problem space and user needs and requirements. By partitioning large problem spaces, like the Problem Space of the Battlefield, into a smaller domain the modeling complexity is greatly reduced. This approach works well with SAs since they are responsible for smaller problem domains. The complexity and efficiency of a system model depends on the number of SAs employed, and the degree of interdependence between them. The larger the number of SAs and the more interdependent they are, the higher the complexity and the lower the efficiency. A supervisory control mechanism must be implemented to insure SA effectiveness when modeling large complex problem spaces. The Virtual Associative Network (VAN) (Yufik Y., U.S. Patent 5586219) is a good candidate for such a supervisory mechanism. This paper offers a rationale for incorporating the VAN as a critical element in the Intelligent Agent (IA) architecture.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jan 01, 2000
- Accession Number
- ADA460984
Entities
People
- Edward Dawidowicz
Organizations
- United States Army Communications-Electronics Command