Analyse des performances d'un système multi-agents par visualisation

The analysis and understanding of computer systems behaviour is an important part in the development cycle. Performance visuali:zation is one of the techniques that can be used to perform a global analysis of a system's behaviour. This technique supports performance debugging. It is essential to identify where and when the ressources are consumed during running time. It pro vides the programmer performance indices and the internai behaviour of the entities in the system. Due to the recent interest in using multi-agent systems in a wide variety of application domains, the analysis and understanding of the internai complex characteristics (interaction, distribution of intelligence, cooperation, coordination, decentralization,. . . ) ofmulti¬agent systems has become a priority. Usually, the behaviour ofa multi-agent system is studied from a global point ofview. The multi-agent system is viewed as a black box where entities interact one with the other in oerder to reach a global goal. Unfortunately, this approach does not provide sufficier information for a complete study ofthe multi-agent system behaviour. A study at the level of the entities composing the multi-agent system is necessar: The performance visualization of a multi-agent system application can be an approach to deal with this problem. Ln this work, we propose an approach to evaluate and to understand the internai behaviour of a multi-agent system application using visualization. We apply this approach on a transport multi-agent application.

[1]  Jiming Liu,et al.  From local behaviors to global performance in a multi-agent system , 2004, Proceedings. IEEE/WIC/ACM International Conference on Intelligent Agent Technology, 2004. (IAT 2004)..

[2]  Jacques Ferber,et al.  Les Systèmes multi-agents: vers une intelligence collective , 1995 .

[3]  Yves Demazeau,et al.  FROM INTERACTIONS TO COLLECTIVE BEHAVIOUR IN AGENT-BASED SYSTEMS , 1995 .

[4]  Jerry C. Yan Performance Tuning with AIMS - An Automated Instrumentation and Monitoring System for Multicomputers , 1994, HICSS.

[5]  D.A. Reed,et al.  Scalable performance analysis: the Pablo performance analysis environment , 1993, Proceedings of Scalable Parallel Libraries Conference.

[6]  Cyril Guilloud Traçage flexible d'exécutions de programmes parallèles. (Flexible trace recording of parallel applications) , 2004 .

[7]  André Schiper,et al.  Execution Replay: A Mechanism for Integrating a Visualization Tool with a Symbolic Debugger , 1992, CONPAR.

[8]  Daniel A. Reed,et al.  Performance Instrumentation Techniques for Parallel Systems , 1993, Performance/SIGMETRICS Tutorials.

[9]  Thomas Meurisse Simulation multi-agent : du modèle à l'opérationnalisation , 2004 .

[10]  Alva L. Couch Categories and Context in Scalable Execution Visualization , 1993, J. Parallel Distributed Comput..

[11]  Ray Jain,et al.  The art of computer systems performance analysis - techniques for experimental design, measurement, simulation, and modeling , 1991, Wiley professional computing.

[12]  Eric Maillet Le traçage logiciel d'applications parallèles : conception et ajustement de qualité. (On the quality of software generated traces of parallel applications) , 1996 .

[13]  Robert E. Shannon,et al.  Introduction to the art and science of simulation , 1998, 1998 Winter Simulation Conference. Proceedings (Cat. No.98CH36274).

[14]  Thomas J. LeBlanc,et al.  Debugging and analysis of large-scale parallel programs , 1989 .

[15]  Yves Arrouye Environnements de visualisation pour l'évaluation des performances des systèmes parallèles : étude, conception et réalisation , 1995 .

[16]  R. Leszczyna,et al.  Evaluation of Agent Platforms , 2004 .