Polymorphic Self-* Agents for Stigmergic Fault Mitigation in Large-Scale Real-Time Embedded Systems

 
EPrints.org
Agentlink Clearinghouse is powered by GNU EPrints developed by the School of Electronics and Computer Science of the University of Southampton.
Type: Conference or Workshop Item (Paper)

Messie, Derek and Oh, Jae C. (2005) Polymorphic Self-* Agents for Stigmergic Fault Mitigation in Large-Scale Real-Time Embedded Systems. In: Fourth International Joint Conference on Autonomous Agents and Multi-Agent Systems (AAMAS), July, 2005, Utrecht, Netherlands.

Full text not available from this archive.

Abstract

Organization and coordination of agents within large-scale, complex, distributed environments is one of the primary challenges in the field of multi-agent systems. A lot of interest has surfaced recently around self-* (self-organizing, self-managing, self-optimizing, self-protecting) agents. This paper presents polymorphic self-* agents that evolve a core set of roles and behavior based on environmental cues. The agents adapt these roles based on the changing demands of the environment, and are directly implementable in computer systems applications. The design combines strategies from game theory, stigmergy, and other biologically inspired models to address fault mitigation in large-scale, real-time, distributed systems. The agents are embedded within the individual digital signal processors of BTeV, a High Energy Physics experiment consisting of 2500 such processors. Results obtained using a SWARM simulation of the BTeV environment demonstrate the polymorphic character of the agents, and show how this design exceeds performance and reliability metrics obtained from comparable centralized, and even traditional decentralized approaches.

Deposited by Derek Messie on 08 August 2005

Archive Staff Only: edit this record

   

AgentLink is the European Commission's IST-funded Coordination Action for Agent-Based Computing
and is coordinated by the
University of Liverpool and University of Southampton
If you encounter any problems with these pages please contact web@agentlink.org.