AQM performance in multiple congested link networks

Authors:
Deepak Vembar;Zhidian Du
Affiliations:
Clemson University, Clemson, SC;Clemson University, Clemson, SC
Venue:
Proceedings of the 43rd annual Southeast regional conference - Volume 2
Year:
2005

Citing 2
Cited 0

TCP and explicit congestion notification

ACM SIGCOMM Computer Communication Review
Tuning RED for Web traffic

Proceedings of the conference on Applications, Technologies, Architectures, and Protocols for Computer Communication

Quantified Score

Hi-index	0.00

Visualization

Abstract

TCP Reno, the most widely used implementation on the Internet, uses retransmission timeouts and the receipt of three duplicate ACKs to detect packet loss in the network. In case of multiple congested links, it is important to detect packet losses as early as possible in order to prevent unnecessary timeouts and retransmits. Active Queue Management (AQM) schemes seek to improve the overall user response time by using queue build-ups at the routers to detect congestion. Previous research has focused on determining the effects of AQM schemes on HTTP/1.0 traffic in single congested link networks [2]. In this experiment, we evaluated the performance of TCP Reno with drop-tail routers and ECN-enabled TCP with Selective Acknowledgements (SACK) [5] over Adaptive Random Early Detection (ARED) [4] routers in a multiple congested link network.