Performance evaluation of congestion window validation for DASH transport

Sajid Nazir, Ziaul Hossain, Raffaello Secchi, Mathew Broadbent, Andreas Pertlund, Gorry Fairhurst

Research output: Chapter in Book/Report/Conference proceedingPublished conference contribution

11 Citations (Scopus)

Abstract

A recent proposed update to TCP congestion control, TCP-newCWV, has targeted congestion control for rate-limited applications. These methods need to be explored in the context of rate-adaptive applications, such as DASH. The new method enables a client to exploit the persistence of a DASH connection and enables the DASH server to rapidly resume transmission of a series of video segments using a single TCP connection. Another technique, called ‘Pacing’ smoothes DASH bursts when there is no TCP ACK clock, and is shown to significantly reduce burst loss. These two methods in combination can increase the application performance. This paper investigates the effect of implementing these techniques on a DASH flow in different congestion scenarios and whether the method can promote better capacity sharing while minimizing the latency experienced by other flows sharing a common network bottleneck. The results confirm that newCWV with Pacing provides a benefit as a platform for DASH transport.
Original languageEnglish
Title of host publicationNOSSDAV '14 Proceedings of Network and Operating System Support on Digital Audio and Video Workshop
Place of PublicationSingapore
PublisherACM Press
Number of pages6
ISBN (Print) 978-1-4503-2706-0
DOIs
Publication statusPublished - 19 Mar 2014
Event24th ACM SIGMM Workshop on Network and Operating Systems Support for Digital Audio and Video (NOSSDAV 2014) - , Singapore
Duration: 19 Mar 201420 Mar 2014

Conference

Conference24th ACM SIGMM Workshop on Network and Operating Systems Support for Digital Audio and Video (NOSSDAV 2014)
Country/TerritorySingapore
Period19/03/1420/03/14

Bibliographical note

NOSDAV

Keywords

  • DASH
  • HAS
  • web video

Fingerprint

Dive into the research topics of 'Performance evaluation of congestion window validation for DASH transport'. Together they form a unique fingerprint.

Cite this