Accuracy of strong and weak comparisons for network of queues

Hind Castel-Taleb 1, * Nihal Pekergin 2
* Auteur correspondant
1 SAMOVAR - Services répartis, Architectures, MOdélisation, Validation, Administration des Réseaux
2 LACL - Laboratoire d'Algorithmique Complexité et Logique
Abstract : Quality of performance measure bounds is crucial for an accurate dimensioning of computer network resources. We study stochastic comparisons of multidimensional Markov processes for which quantitative analysis could be intractable if there is no specific solution form. On partially ordered state space, different stochastic orderings can be defined as the strong or the less constrained weak ordering. The goal of the present paper is to compare these two orderings with respect the quality of derived bounds. We propose to study a system similar to a Jackson network except that queues have finite capacity. Different bounding systems are built either in the sense of the strong ordering with hard constraints, or in the sense of the weak ordering with less ones. The proofs of the stochastic comparisons are done using the coupling and the increasing set methods, with an intuitive event based formalism. The qualities of bounding systems are compared regarding to blocking probabilities.
Type de document :
Communication dans un congrès
15th International GI/ITG Conference, MMB&DFT 2010, Mar 2010, Essen, Germany. Springer Verlag, 5987, www.springerlink.com, 2010, 〈10.1007/978-3-642-12104-3_17〉
Liste complète des métadonnées
https://hal-imt.archives-ouvertes.fr/hal-00545740
Contributeur : Hind Castel <>
Soumis le : samedi 11 décembre 2010 - 19:31:41
Dernière modification le : jeudi 3 mai 2018 - 16:58:02

Identifiants

Collections

LACL | TELECOM-SUDPARIS | UPEC | UPEC-UPEM

Citation

Hind Castel-Taleb, Nihal Pekergin. Accuracy of strong and weak comparisons for network of queues. 15th International GI/ITG Conference, MMB&DFT 2010, Mar 2010, Essen, Germany. Springer Verlag, 5987, www.springerlink.com, 2010, 〈10.1007/978-3-642-12104-3_17〉. 〈hal-00545740〉

Exporter

BibTeX TEI DC DCterms EndNote

Partager

Métriques

Consultations de la notice

122

Contact

support.ccsd.cnrs.fr
hal.support@ccsd.cnrs.fr
Logo CCSD