doi:10.1016/j.peva.2004.08.005
Copyright © 2004 Elsevier B.V. All rights reserved.
Routing and rate allocation in rate-based multi-class networks
References and further reading may be available for this article. To view references and further reading you must
purchase this article.
Aniruddha Diwana, 
, 
and Joy Kurib,
aMotorola India Electronics Pvt. Ltd., “The Senate”, Ulsoor Road, Bangalore 560042, India
bCentre for Electronics Design and Technology, Indian Institute of Science, Bangalore 560012, India
Received 6 December 2003;
revised 23 August 2004.
Available online 8 October 2004.
Abstract
In this paper, we investigate the performance of routing and rate allocation (RRA) algorithms in rate-based multi-class networks. On the arrival of a connection request, an RRA algorithm selects a route for the connection and allocates an appropriate rate on the route; failing this, it blocks the connection request. We measure the performance of an RRA algorithm in terms of its minimum weighted carried traffic. This performance criterion encompasses two widely used performance criteria, namely, weighted carried traffic and minimum carried traffic. We derive an upper bound on the minimum weighted carried traffic of any RRA algorithm. The bound can be computed by solving a linear program. Moreover, we show that the bound is achieved asymptotically, when the offered load and the link capacities are large, by a Partitioning RRA algorithm. Therefore the bound can be used as a performance benchmark for any RRA algorithm. We observe that the proposed Partitioning RRA algorithm, though asymptotically optimal, performs poorly at very low loads. We investigate the cause of this undesirable behaviour and obtain two improved asymptotically optimal RRA algorithms.
Keywords: Routing; Rate allocation; Multi-class networks
Fig. 1. Network to illustrate the performance of the SP-RRA scheme. It has 14 nodes and 21 links. The link rates are 155 Mbps and the link propagation delays are 4 ms.
Fig. 2. NSFNET: uniform traffic: The LP upper bound and the weighted carried traffics obtained from the Partitioning RRA and the SP-RRA in the low blocking region: (top) scale factor 1 (middle) scale factor 5 (bottom) scale factor 20. The performance of Partitioning RRA is poor at scale factor 1. The Partitioning RRA is optimal only asymptotically. The SP-RRA performs well at very low load values.
Fig. 3. NSFNET: nonuniform traffic: The LP upper bound and the weighted carried traffics obtained from the Partitioning RRA and the SP-RRA in the low blocking region: (top) scale factor 1 (middle) scale factor 5 (bottom) scale factor 20. The performance of Partitioning RRA is poor at scale factor 1. The Partitioning RRA is optimal only asymptotically. The SP-RRA performs well at very low load values.
Fig. 4. NSFNET: uniform traffic: The LP upper bound and the weighted carried traffic obtained from the Partitioning RRA and the SP-RRA: full range: (top) scale factor 1 (middle) scale factor 5 (bottom) scale factor 20.
Fig. 5. NSFNET: nonuniform traffic: The LP upper bound and the weighted carried traffic obtained from the Partitioning RRA and the SP-RRA: full range: (top) scale factor 1 (middle) scale factor 5 (bottom) scale factor 20.
Fig. 6. NSFNET: uniform traffic: The LP upper bound and the minimum carried traffics obtained from the Partitioning RRA and the SP-RRA in the low blocking region: (top) scale factor 1 (middle) scale factor 5 (bottom) scale factor 20. The performance of Partitioning RRA is poor at scale factor 1. The Partitioning RRA is optimal only asymptotically. The SP-RRA performs well at very low load values.
Fig. 7. NSFNET: nonuniform traffic: The LP upper bound and the minimum carried traffics obtained from the Partitioning RRA and the SP-RRA in the low blocking region: (top) scale factor 1 (middle) scale factor 5 (bottom) scale factor 20. The performance of Partitioning RRA is poor at scale factor 1. The Partitioning RRA is optimal only asymptotically. The SP-RRA performs well at very low load values.
Fig. 8. NSFNET: uniform traffic: The LP upper bound and the minimum carried traffic obtained from the Partitioning RRA and the SP-RRA: full range: (top) scale factor 1 (middle) scale factor 5 (bottom) scale factor 20.
Fig. 9. NSFNET: nonuniform traffic: The LP upper bound and the minimum carried traffic obtained from the Partitioning RRA and the SP-RRA: full range: (top) scale factor 1 (middle) scale factor 5 (bottom) scale factor 20.
Fig. 10. NSFNET: uniform traffic: The LP upper bound and the performances of the Hybrid RRA, the Improved Partitioning RRA and the Partitioning RRA for the low blocking region: (top) scale factor =1 (middle) scale factor 5 (bottom) scale factor 20. The Improved Partitioning RRA and the Hybrid RRA policies perform well even at very low load values. Overall, the Hybrid RRA performs better than the rest.
Table 1.
Parameters for traffic models of type voice and video
Parts of this work appeared in Proceedings of IEEE GLOBECOM 2002 and Proceedings of NCC 2003.
Corresponding author.
Abstract
Purchase PDF (408 K)
E-mail Article
Add to my Quick Links
Cited By in Scopus (0)
Purchase PDF (1561 K)
