Title
Performance evaluation framework for IP resource reservation signalling
Abstract
This paper investigates the scalability limitations of signalling-based Quality of Service (QoS) provisioning using the RSVP and Boomerang protocols as examples. We establish a performance evaluation framework, which contains per hop metrics for describing the router's signalling capability and per reservation metrics for characterising the overall performance of the reservation protocol. The per hop metrics are captured in a simple router model, using parameters that are based on actual measurements. This router model is used in a simulation scenario, which characterises the end-to-end impact and overall performance of the RSVP and Boomerang reservation protocols.This paper summarises the results of router benchmarking measurements and simulations. The per session results highlight that, with thorough selection of the control protocol, signalling-based resource reservation is a realistic approach in the Internet. It is also shown that the selection of the signalling protocol can seriously influence the reservation performance of a router and thus the overall performance of the entire network as well. The presented performance evaluation framework can be used for network dimensioning and it can even aid the design of new resource reservation protocols and new router devices.
Year
DOI
Venue
2002
10.1016/S0166-5316(02)00034-2
Perform. Eval.
Keywords
Field
DocType
performance evaluation framework,ip quality of service,reservation performance,rsvp,simulation,signalling-based resource reservation,reservation metrics,benchmarking,overall performance,router benchmarking measurement,new router device,new resource reservation protocol,hop metrics,ip resource reservation signalling,boomerang reservation protocol,reservation protocol,quality of service
Resource Reservation Protocol,Reservation,Computer science,Quality of service,Computer network,Provisioning,Real-time computing,Router,Benchmarking,The Internet,Scalability,Distributed computing
Journal
Volume
Issue
ISSN
48
1-4
Performance Evaluation
Citations 
PageRank 
References 
1
0.36
8
Authors
3
Name
Order
Citations
PageRank
Gábor Fehér1276.03
Krisztián Németh2173.29
István Cselényi3142.67