论文信息 - Quantitative Analysis of Single-Level Single-Mediator Multi-agent Systems

Quantitative Analysis of Single-Level Single-Mediator Multi-agent Systems

Queueing Theory deals with problems where some restricted resource should be shared between competitive flow of requests. In this paper we use Queueing Theory methods to perform a quantitative analysis of a single-level single-mediator multi-agent system. In the system, several agents, coordinated by the mediator process user queries. We adopt matrix analytic methods to compute performance characteristics in terms of a queueing network of tree-like topology with cooperation of the servers. Results can be used for the logical and technical design and optimal resources sharing in multi-agent systems.

Alexander N. Dudin | Moon Ho Lee | Aliaksandr Birukou | Valentina I. Klimenok | Chang-hui Choe

[1] Steven Okamoto,et al. Toward an understanding of the impact of software personal assistants on human organizations , 2006, AAMAS '06.

[2] Victor R. Lesser,et al. Using queuing theory to predict organizational metrics , 2006, AAMAS '06.

[3] Valentina Klimenok,et al. The SM/M/N queueing system with broadcasting service , 2006 .

[4] Nathan Gnanasambandam. Survivability of multi-agent systems , 2005, AAMAS '05.

[5] Victor R. Lesser,et al. A survey of multi-agent organizational paradigms , 2004, The Knowledge Engineering Review.

[6] Javier Vázquez-Salceda,et al. Organizing Multiagent Systems , 2005, Autonomous Agents and Multi-Agent Systems.

[7] Victor Lesser,et al. A dynamically formed hierarchical agent organization for a distributed content sharing system , 2004 .

[8] M. Brinn,et al. Reliable MAS performance prediction using queueing models , 2004, IEEE First Symposium onMulti-Agent Security and Survivability, 2004.