Models and analysis of TCC/AQM schemes over diffserv networks

Authors:
Jahwan Koo;Jitae Shin;Seongjin Ahn;Jinwook Chung
Affiliations:
School of Information and Communications Engineering, Sungkyunkwan Univ., Kyounggi-do, Korea;School of Information and Communications Engineering, Sungkyunkwan Univ., Kyounggi-do, Korea;Department of Computer Education, Sungkyunkwan Univ., Seoul, Korea;School of Information and Communications Engineering, Sungkyunkwan Univ., Kyounggi-do, Korea
Venue:
ICN'05 Proceedings of the 4th international conference on Networking - Volume Part I
Year:
2005

Citing 3
Cited 1

Congestion avoidance and control

SIGCOMM '88 Symposium proceedings on Communications architectures and protocols
Random early detection gateways for congestion avoidance

IEEE/ACM Transactions on Networking (TON)
TCP Vegas: end to end congestion avoidance on a global Internet

IEEE Journal on Selected Areas in Communications

An extended framework for proportional differentiation: performance metrics and evaluation considerations

ICCSA'05 Proceedings of the 2005 international conference on Computational Science and Its Applications - Volume Part II

Quantified Score

Hi-index	0.00

Visualization

Abstract

QoS-enabled networks consist of major functions: one is a TCP congestion control (TCC) mechanism at a source/sink node and the other is an active queue management (AQM) scheme at an intermediate node. In this paper, we introduce the major TCC mechanisms and AQM schemes, and provide analytical models and simulation-based comparisons of these TCC/AQM schemes for the purpose of identifying the reciprocal relationship between TCC mechanisms and AQM schemes in QoS-enabled networks. The results show that the equilibrium and dynamics of the underlying network depends on the harmony between the TCC/AQM pairs. The NewReno/PI pair, a feedback-based mechanism encompassing both network and end-systems, can enhance the performance of packet loss and delay sensitive applications. In our opinion, an appropriate combination of active queue management from the network and TCP source reaction would provide an effective solution to the instantaneous network fluctuation which occurs on the Internet.