Autonomous Agents for Digital Network Maximization
Abstract
The advent of the computer age has brought about a plenitude of benefits to the human race. Included with these benefits has been the ever increasing demand to transfer exponentially increasing amounts of information, and the associated problems of information sharing. The focus of this thesis, Naval Science Assistance Program (NSAP). and Office of Naval Research (ONR) funded research effort, has been to best utilize available digital communications assets in the radio frequency (RF) spectrum to allow sufficient transfer of information providing DOD assets flexible, rapid, and in-flight reprogramming, re-planning of strike and cruise missile assets, to engage a high value, emergent target, in the shortest possible time. The postulated methods of utilizing autonomous agents to manage information flow across network nodes has applicability to all digital networks. Based. upon the pioneering work of Pattie Maes, at Massachusetts Institute of Technology (MET), and previous examination of communications node management, the implementation of independent processes, working on behalf of a host system, to optimize the effective meaningful throughput on a communications channel is not only desirable, but necessary. The evolution of semi intelligent software, whether called Artificial Intelligence, Intelligent Agents, or Autonomous Agents, has reached a level of sophistication allowing the insertion of meaningful articulated processes within existing, and future systems to maximize the network efficiency systematically. Recent work by Michael Cohen on sodabots, and the evolution of user interactive TinyMUDS of the Maas-Neotek family, a virtual type personality environment, has demonstrated the ability of software to deal with dynamic and changing conditions.
Document Details
- Document Type
- Technical Report
- Publication Date
- Sep 01, 1997
- Accession Number
- ADA337729
Entities
People
- Michael W. Dabose
Organizations
- Naval Postgraduate School