arrow
Return

RCFT: A Termination Method for Simple PCN-Based Flow Control

delete2013-01-30
delete1
PRE
AI
F
Frank Lehrieder
M
Michael Menth *
DOI:10.1007/s10922-013-9264-6delete
deleteOriginal
deleteOriginal request for help
deleteShare
deleteSave
Abstract

Abstract

En 中文
Pre-congestion notification (PCN) conveys information about load conditions in differentiated services IP networks to boundary nodes. This information is currently used for admission control and flow termination. Flow termination complements admission control, e.g., in case of failures when admitted traffic is rerouted and causes overload on backup paths. Existing approaches for PCN-based admission control and flow termination operate on ingress-egress aggregates and rely on a signalling protocol that regularly reports measured PCN feedback from all egress nodes to all ingress nodes. However, this signalling protocol is neither defined nor available, and the methods have also other intrinsic shortcomings that result from their operations on ingress-egress aggregates. While there is already a PCN-based admission control method that works without additional signalling of measured PCN feedback, a solid flow termination method with that property is still missing. In this paper we present the novel regular-check-based flow termination method (RCFT). It does not rely on measured PCN feedback, fills the identified gap, and allows for a PCN architecture without signalling of measured feedback. We explain RCFT in detail and investigate its termination behavior under various conditions. Moreover, we study the use of PCN-based flow control for on/off traffic. These results are of general nature and apply to any system using PCN-based flow termination.
Keywords:
Quality of service
Pre-congestion notification
Admission control
Flow termination
AI Summary

AI Summary

Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.

Journal

Journal of Network and Systems Management cover
Journal of Network and Systems Management
IF:
3.9
Papers:
1.0K
Citations:
1.3K

Organization

E
eberhard karls university of tubingen
Scholars:
3.3W
Papers: 2.5W
Citations: 38
U
University of Wurzburg
Scholars:
2.5W
Papers: 2.0W
Citations: 2.5W