Retour au site du LAAS-CNRS

Mots-Clés / Keywords

Abstract

Network experiments are essential for assessing and validating new networking technologies, architectures and protocols. These assessments have long been performed using network simulators. But it clearly appeared that the results got in simulations cannot be reproduced in real environment. Emulators can hardly integrate accurate models of all networking components, end host operating systems and applications what leads to unrealistic simulations very often. Therefore, some work has been issued for developing real experiment platform and network emulators. This paper addresses the motivations that raised the design and development of such an experimental platform at LAAS ⬠laasnetexp ⬠and describes its constituting features and components. It is in particular detailed how experimental conditions can be fully controlled for reproducible and easy to analyze experiments. Last, this paper describes how realistic conditions can be set-up in experiments by using the results of actual Internet and Internet traffic characterization, analysis and modeling. Such information helps to realistically configure emulators as well as define realistic traffic generators. The realism of such experiments is illustrated as a demonstration of the interest of laasnetexp for networking research.

Mots-Clés / Keywords

Abstract

We present a fixed point approach to evaluate the quality of service of streaming traffic multiplexed with elastic traffic in multi-service network. First, we handle elastic traffic66 and streaming traffic separately, and then we derive a general fixed point formulation integrating both types of traffic in best effort networks. Then, we extend the application of this formulation to multi-service networks where priorities and bandwidth sharing schemes can be applied to different flows. Our approach is mainly oriented towards very large scale networks where traditional simulation techniques are not scalable, and where a large number of flows have to be evaluated in reasonable time. We assess the accuracy of our approach by means of event-driven simulations.

Mots-Clés / Keywords

Résumé

L'expérimentation en réseaux est incontournable pour évaluer et valider de nouvelles technologies, architectures et protocoles de communication. Il s'avère que les résultats obtenus à partir de simulations sont la plupart du temps irréalistes, les simulateurs ne pouvant pas intégrer tous les modèles de comportement des composants réseaux, des systèmes opératoires et des applications des machines d'extrémité. Des travaux autour de l'émulation et de l'expérimentation réelle en réseaux ont donc vu le jour. Ce papier présente les motivations et besoins qui ont conduit à la mise en place d'une telle plate-forme expérimentale au LAAS - LaasNetExp - et décrit ses principes de conception. L'article détaille comment les conditions expérimentales peuvent être contrôlées pour des expériences reproductibles et faciles à analyser. L'article montre aussi comment des conditions expérimentales réalistes configuration des émulateurs et générateurs de trafics) peuvent être mise en oeuvre à partir des résultats de caractérisation et analyse du trafic Internet. Cet article évalue le réalisme des expérimentations ainsi obtenues.

Mots-Clés / Keywords