Games and Queues

Bernardo A. Huberman, Li Zhang and Fang Wu

No 336, Computing in Economics and Finance 2004 from Society for Computational Economics

Abstract: We consider scheduling in distributed systems from a game theoretic point view while taking into account queuing theory methodologies. In this approach no one knows the global state of the system while users try to maximize their utility. Since the performance of such a blind scheduler is worse than the optimal, it induces users to employ strategies to improve their own utilization of the system. One such strategy is that of restarting a request if it is not satisfied in a given time. Since we assume users as non-cooperative and selfish, the problem is that of studying the characteristic of the Nash equilibrium in a large distributed system with no omniscient controls. We study the problem through computer experiments and analytical approaches. We obtain exact solutions in situations delimited by two extremes: one in which users never restart an initial request, and another one in which the user's requests are restarted infinitely often. Users can switch between these two behaviors. When the system load is below certain threshold, it is always better off to be impatient, and when the system load is higher than some threshold, it is always better to be patient. Between them there exists a homogeneous Nash equilibrium with non-trivial properties.

Keywords: distributed; systems (search for similar items in EconPapers)
JEL-codes: C7 (search for similar items in EconPapers)
Date: 2004-08-11
References: Add references at CitEc
Citations:

There are no downloads for this item, see the EconPapers FAQ for hints about obtaining it.

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:sce:scecf4:336

Access Statistics for this paper

More papers in Computing in Economics and Finance 2004 from Society for Computational Economics Contact information at EDIRC.
Bibliographic data for series maintained by Christopher F. Baum ().