Şebnem BORA, Ali Murat TİRYAKİ, Oğuz DİKENELLİ

Load sharing based on moving roles in multiagent systems

In this paper, we present a load-sharing approach based on the refactoring of agents. According to our approach, the role(s) that makes an agent overloaded is identified and transferred to less loaded agents. The excess workload of this heavily loaded agent is then transferred to the new agent. This approach defines a new agent, called the ``monitor agent,'' which monitors the workload of agents in the organization and decides about the refactoring of the agents. The monitor agent uses the platform ontology, which explicitly describes the components of the agent organization, including agents and their roles, plans, and workloads. This ontology is updated by the monitor agent in every monitor cycle.

Anahtar Kelimeler:

Distributed computing, dynamic load sharing, multiagent systems, software agents

Load sharing based on moving roles in multiagent systems

Keywords:

Distributed computing, dynamic load sharing, multiagent systems, software agents,

PDF

___

M. Fowler, Refactoring: Improving the Design of Existing Code, Boston, Addison Wesley, 2000.
A. Schaerf, Y. Shoham, M. Tennenholtz, “Adaptive load balancing: a study in multi-agent learning”, Journal of ArtiŞcial Intelligence Research, Vol. 2, pp. 475-500, 1995.
S. Appleby, S. Steward, “Mobile software agents for control in telecommunications networks”, BT Technology Journal, Vol. 18, pp. 68-70, 2000.
J. Cao, D. Spooner, S. Jarvis, G. Nudd, “Grid load balancing using intelligent agents”, Future Generation Computer Systems, Vol. 21, pp. 135-149, 2005.
M. Singhal, N.G. Shivaratri, Advanced Concepts in Operating Systems, New York, McGraw-Hill, 1994.
F. Giunchiglia, J. Mylopoulos, A. Perini, “The Tropos software development methodology: processes, models, and diagrams”, Proceedings of First International Conference on Autonomous Agents and Multiagent Systems, pp. 36, 2002.
F. Zambonelli, N.R. Jennings, M. Wooldridge, “Developing multiagent systems: the Gaia methodology”, ACM Transactions on Software Engineering and Methodology, Vol. 12, pp. 317-370, 2003.
FIPA, Review of FIPA SpeciŞcations, available at http://www.Şpa.org/subgroups/ROFS-SG-docs/ROFS-Doc.pdf, Web-Ontology Working Group, Conclusions and Future Work, available at http://www.w3.org/2001/sw/WebOnt/,
S. Dhakal, M.M. Hayat, J.E. Pezoa, C. Yang, D.A. Bader, “Dynamic load balancing in distributed systems in the presence of delays: a regeneration-theory approach”, IEEE Transactions on Parallel & Distributed Systems, Vol. , pp. 485-497, 2007.
A. Papoulis, Probability, Random Variables, and Stochastic Processes, 3rd ed., New York, McGraw-Hill, pp. 648- , 1991.
O. Dikenelli, R.C. Erdur, O. Gumus, “SEAGENT: A platform for developing semantic web based multi agent systems”, Fourth International Joint Conference on Autonomous Agents, pp. 1271-1272, 2005.
M. Paolucci, O. Shehory, K. Sycara, D. Kalp, A. Pannu, “A planning component for RETSINA agents”, Lecture Notes in Computer Science, Vol. 1757/2000, pp. 147-161, 2000.
M. Williamson, K. Decker, K. Sycara, “UniŞed information and control ﬂow in hierarchical task networks”, Working Notes of the AAAI-96 Workshop “Theories of Action, Planning, and Control”, 1996. Ege University, SEAGENT: A Semantic Web Enabled Multi-Agent System Framework, available at http://www.seagent.ege.edu.tr, 2006.