Evaluation of Priority Scheduling and Flow Starvation for Thin Streams with FQ-CoDel

Eduard Grigorescu, Chamil Premantha Weerasinghe Kulatunga, Gorry Fairhurst, Nicolas Kuhn

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Bufferbloat is the result of oversized buffers and induced high end- to-end latency experienced by applications across the Internet. This additional delay can adversely impact thin streams, that frequently exchange small amounts of data, but have stringent latency requirements. Active Queue Management (AQM) techniques, such as Controlled Delay (CoDel), can control the queuing delay in a network device to ensure low latency by dropping packets to indicate incipient congestion. FlowQueue-CoDel (FQ-CoDel) is a scheduling scheme that creates one sub-queue per flow and apply CoDel on each of them. FQ-CoDel features: (1) priority scheduling for low-rate traffic; (2) flow isolation; (3) queue management with CoDel. First, this paper fills a gap in the understanding of FQ- CoDel by analyzing what features are of interests for providing low latency for thin streams applications. Second, this paper provides the first analysis of the limits of the flow starvation mechanisms and show that FQ-CoDel is vulnerable to Denial of Service (DoS) attacks.
Original languageEnglish
Title of host publication2015 European Conference on Networks and Communications
Subtitle of host publicationTrack 7: Special Sessions
Number of pages6
Publication statusPublished - 29 Jun 2015

Fingerprint

Scheduling
Internet
Denial-of-service attack

Keywords

  • CoDel
  • AQM

Cite this

Grigorescu, E., Kulatunga, C. P. W., Fairhurst, G., & Kuhn, N. (2015). Evaluation of Priority Scheduling and Flow Starvation for Thin Streams with FQ-CoDel. In 2015 European Conference on Networks and Communications: Track 7: Special Sessions

Evaluation of Priority Scheduling and Flow Starvation for Thin Streams with FQ-CoDel. / Grigorescu, Eduard; Kulatunga, Chamil Premantha Weerasinghe; Fairhurst, Gorry; Kuhn, Nicolas.

2015 European Conference on Networks and Communications: Track 7: Special Sessions . 2015.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Grigorescu, E, Kulatunga, CPW, Fairhurst, G & Kuhn, N 2015, Evaluation of Priority Scheduling and Flow Starvation for Thin Streams with FQ-CoDel. in 2015 European Conference on Networks and Communications: Track 7: Special Sessions .
Grigorescu E, Kulatunga CPW, Fairhurst G, Kuhn N. Evaluation of Priority Scheduling and Flow Starvation for Thin Streams with FQ-CoDel. In 2015 European Conference on Networks and Communications: Track 7: Special Sessions . 2015
Grigorescu, Eduard ; Kulatunga, Chamil Premantha Weerasinghe ; Fairhurst, Gorry ; Kuhn, Nicolas. / Evaluation of Priority Scheduling and Flow Starvation for Thin Streams with FQ-CoDel. 2015 European Conference on Networks and Communications: Track 7: Special Sessions . 2015.
@inproceedings{55fc27e2d6de44c4b373572fa8dd850f,
title = "Evaluation of Priority Scheduling and Flow Starvation for Thin Streams with FQ-CoDel",
abstract = "Bufferbloat is the result of oversized buffers and induced high end- to-end latency experienced by applications across the Internet. This additional delay can adversely impact thin streams, that frequently exchange small amounts of data, but have stringent latency requirements. Active Queue Management (AQM) techniques, such as Controlled Delay (CoDel), can control the queuing delay in a network device to ensure low latency by dropping packets to indicate incipient congestion. FlowQueue-CoDel (FQ-CoDel) is a scheduling scheme that creates one sub-queue per flow and apply CoDel on each of them. FQ-CoDel features: (1) priority scheduling for low-rate traffic; (2) flow isolation; (3) queue management with CoDel. First, this paper fills a gap in the understanding of FQ- CoDel by analyzing what features are of interests for providing low latency for thin streams applications. Second, this paper provides the first analysis of the limits of the flow starvation mechanisms and show that FQ-CoDel is vulnerable to Denial of Service (DoS) attacks.",
keywords = "CoDel, AQM",
author = "Eduard Grigorescu and Kulatunga, {Chamil Premantha Weerasinghe} and Gorry Fairhurst and Nicolas Kuhn",
year = "2015",
month = "6",
day = "29",
language = "English",
booktitle = "2015 European Conference on Networks and Communications",

}

TY - GEN

T1 - Evaluation of Priority Scheduling and Flow Starvation for Thin Streams with FQ-CoDel

AU - Grigorescu, Eduard

AU - Kulatunga, Chamil Premantha Weerasinghe

AU - Fairhurst, Gorry

AU - Kuhn, Nicolas

PY - 2015/6/29

Y1 - 2015/6/29

N2 - Bufferbloat is the result of oversized buffers and induced high end- to-end latency experienced by applications across the Internet. This additional delay can adversely impact thin streams, that frequently exchange small amounts of data, but have stringent latency requirements. Active Queue Management (AQM) techniques, such as Controlled Delay (CoDel), can control the queuing delay in a network device to ensure low latency by dropping packets to indicate incipient congestion. FlowQueue-CoDel (FQ-CoDel) is a scheduling scheme that creates one sub-queue per flow and apply CoDel on each of them. FQ-CoDel features: (1) priority scheduling for low-rate traffic; (2) flow isolation; (3) queue management with CoDel. First, this paper fills a gap in the understanding of FQ- CoDel by analyzing what features are of interests for providing low latency for thin streams applications. Second, this paper provides the first analysis of the limits of the flow starvation mechanisms and show that FQ-CoDel is vulnerable to Denial of Service (DoS) attacks.

AB - Bufferbloat is the result of oversized buffers and induced high end- to-end latency experienced by applications across the Internet. This additional delay can adversely impact thin streams, that frequently exchange small amounts of data, but have stringent latency requirements. Active Queue Management (AQM) techniques, such as Controlled Delay (CoDel), can control the queuing delay in a network device to ensure low latency by dropping packets to indicate incipient congestion. FlowQueue-CoDel (FQ-CoDel) is a scheduling scheme that creates one sub-queue per flow and apply CoDel on each of them. FQ-CoDel features: (1) priority scheduling for low-rate traffic; (2) flow isolation; (3) queue management with CoDel. First, this paper fills a gap in the understanding of FQ- CoDel by analyzing what features are of interests for providing low latency for thin streams applications. Second, this paper provides the first analysis of the limits of the flow starvation mechanisms and show that FQ-CoDel is vulnerable to Denial of Service (DoS) attacks.

KW - CoDel

KW - AQM

M3 - Conference contribution

BT - 2015 European Conference on Networks and Communications

ER -