Performance of Traffic Aggregates under Guaranteed-Rate Scheduling Algorithms

Project Description

Real-time applications, such as voice-over-IP (VoIP? ) and video conferencing, require the network to provide better Quality-of-Service (QoS) in terms of delay, jitter, and loss rate. To provide such QoS support, several network architectures have been proposed, including IntServ? and DiffServ. To better balance performance and scalability, we consider an alternative by extending the IntServ? architecture to support traffic aggregation. This extension is made on the premise that there are aggregation regions in the network, which "see" only aggregated (not individual) flows. Resource reservation and packet scheduling in an aggregation region are done on a per-aggregate basis. This project focuses on the scheduling issues associated with traffic aggregation. It evaluates the end-to-end delay of aggregate scheduling when guaranteed-rate algorithms such as Weighted Fair Queueing (WFQ) are used.

People

Faculty

• Kang G. Shin

Student

• Wei Sun

Publications

• Wei Sun and Kang G. Shin. End-to-End Delay Bounds for Traffic Aggregates under Guaranteed-Rate Scheduling Algorithms. IEEE/ACM Transactions on Networking. , Vol. 13, No. 5, pp. 1188-1201, October, 2005. PDF.
• Wei Sun and Kang G. Shin. TCP Performance under Aggregate Fair Queueing. Proceedings of IEEE GLOBECOM 2004. , pp. 1308-1313, November, 2004. PDF.
• Wei Sun and Kang G. Shin. Coordinated Aggregate Scheduling for Improving End-to-End Delay Performance. Proceedings of 12th IEEE International Workshop on Quality of Service (IWQoS? '04), Montreal, Canada. , June, 2004. PDF.
• Wei Sun and Kang G. Shin. End-to-End Delay Bounds for Traffic Aggregates under Guaranteed-Rate Scheduling Algorithms. Department of EECS, University of Michigan, Technical Report CSE-TR-484-03. , 2003.

-- Main.wsunz - 21 Mar 2007