Home | Issues | Profile | History | Submission | Review
Vol: 52(66) No: 1 / March 2007      

Packet and Queue Size Variability and the Accuracy of Analytical Models for DiffServ Networks
Stejarel Veres
School of Information Technology and Engineering, University of Ottawa, 161 Louis Pasteur, Ottawa, ON K1N 6N5, Canada, phone: (613) 562-5800, e-mail: cveres@site.uottawa.ca, web: http://www.ncct.uottawa.ca
Dan Ionescu
School of Information Technology and Engineering, University of Ottawa, 161 Louis Pasteur, Ottawa, ON K1N 6N5, Canada, phone: (613) 562-6209, e-mail: ionescu@site.uottawa.ca, web: http://www.ncct.uottawa.ca


Keywords: Quality of Service, Differentiated Services, packet size variability, model accuracy, interpolation.

Abstract
In the past few years, Quality of Service has become a must for IP-based internetworks due to the fact that some of their latest applications, such as Voice and Video over IP, IP telephony, multimedia collaboration, or tele-medicine, are extremely sensitive to QoS parameters. Since IP has no inherent support for QoS guarantees, different architectures that aim to enhance IP in that regard have been devised. The two most prominent such architectures are Integrated Services (IntServ) and Differentiated Services (DiffServ). Of these two, DiffServ seems to have had more impact in the networking market. It has been shown, however, that current DiffServ implementations do not follow a common, unified mathematical model and therefore differ in performance, subject to vendor-specific hardware and software architectures and optimizations. In order to address this, several models have been proposed, but are widely regarded as work in progress. This paper examines some assumptions that have been made throughout the definition of such models so far, and tries to experimentally analyze how these assumptions affect the model accuracy. Several shortcomings of current models are identified, and the proposed approach to address these issues is outlined. This is part of a larger work that aims to propose a comprehensive and accurate model for DiffServ-aware QoS implementations.

References
[1] B. Ionescu, M. Ionescu, S. Veres, and D. Ionescu, “Design and Architecture of an IP over Optical and Optical Ethernet Next Generation Network – NCIT*net 2,” in Proceedings of the 4th International Information and Telecommunication Technologies Symposium (I2TS’2005), Florianópolis, Brazil, December 2005.
[2] R. Braden, D. Clark, and S. Shenker, “Integrated Services in the Internet Architecture: and Overview,” Internet Engineering Task Force, Sterling, VA, RFC 1633, June 1994.
[3] S. Blake, D. Black, M. Carlson, E. Davies, Z. Wang, and W. Weiss, “An Architecture for Differentiated Service,” Internet Engineering Task Force, Sterling, VA, RFC 2475, December 1998.
[4] J. Heinanen, F. Baker, W. Weiss, and J. Wroclawski, “Assured Forwarding PHB Group,” Internet Engineering Task Force, Sterling, VA, RFC 2597, June 1999.
[5] B. Davie, A. Charny, J. C. R. Bennet, K. Benson, J. Y. Le Boudec, W. Courtney, S. Davari, V. Firoiu, and D. Stiliadis, “An Expedited Forwarding PHB (Per-Hop Behaviour),” Internet Engineering Task Force, Sterling, VA, RFC 3246, March 2002.
[6] S. Floyd and V. Jacobson, “Random Early Detection Gateways for Congestion Avoidance,” IEEE/ACM Transactions on Networking, vol. 1, no. 4, pp. 397–413, August 1993.
[7] H. M. Alazemi, A. Mokhtar, and M. Azizoglu, “Stochastic Approach for Modeling Random Early Detection Gateways in TCP/IP Networks,” in Proceedings of IEEE International Conference on Communications 2001 (ICC 2001), vol. 8, June 2001, pp. 2385–2390.
[8] T. Bonald, M. May, and J.-C. Bolot, “Analytic Evaluation of RED Performance,” in Proceedings of Nineteenth Annual Joint Conference of the IEEE Computer and Communications Societies (INFOCOM 2000), 2000, pp. 1415–1424.
[9] H. Su and M. Atiquzzaman, “Comprehensive Performance Model of Differentiated Service with Token Bucket Marker,” in IEE Proceedings in Communications, vol. 150, no. 5, October 2003, pp. 347–353.
[10] S. Veres and D. Ionescu, “A Per-Flow Measurement and Estimation Method for DiffServ Implementations”, in Proceedings of the IEEE Instrumentation and Measurement Technology Conference (IMTC 2006), Sorrento, Italy, April 24-27, 2006, pp. 353–357, ISBN 0-7803-9360-0.
[11] S. Veres and D. Ionescu, “A Performance Model and Measurement Framework for DiffServ Implementations”, submitted to IEEE Transactions on Instrumentation and Measurement, IMTC 2006 Special Edition.
[12] Eric W. Weisstein. Standard Deviation. From MathWorld – A Wolfram Web Resource, http://mathworld.wolfram.com/ StandardDeviation.html.
[13] S. Veres and D. Ionescu, “A Measurement-Oriented Approach to Modeling Packet Loss in IP Networks”, in Proceedings of the IEEE Instrumentation and Measurement Technology Conference (IMTC 2007), Warsaw, Poland, May 1-3, 2007 (in press).
[14] D. W. Harder. Numerical Methods and Analysis for Engineers, http://www.ece.uwaterloo.ca/~ece204/TheBook.