Adaptive redundancy scheme for scalable video traffic using network coding

Hana Baccouch, Paul Louis Ageneau, Nadia Boukhatem

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

In this paper, we introduce a new network coding scheme with adaptive redundancy for scalable video streams in multi-hop wireless networks. Since scalable video coding produces a high quality bitstream with various levels of quality, we propose to offer better protection for important video layers when the link experiences a poor quality. Our scheme adapts the amount of redundancy depending on the quality of the output link and packet's importance. Our approach uses nested windows inspired by unequal protection schemes for scalable video coding. We carried out a simulation based evaluation of our scheme considering packet loss and overhead metrics in various network scenarios. The results show that our scheme reduces the overall packet loss and offers a better protection for the base layer packets.

Original languageEnglish
Title of host publication2016 7th International Conference on the Network of the Future, NOF 2016
EditorsOtto Carlos Muniz Bandeira Duarte, Stefano Secci, Marcelo Goncalves Rubinstein, Marcelo Dias de Amorim
PublisherInstitute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)9781509046713
DOIs
Publication statusPublished - 6 Jan 2017
Externally publishedYes
Event2016 7th International Conference on the Network of the Future, NOF 2016 - Buzios, RJ, Brazil
Duration: 16 Nov 201618 Nov 2016

Publication series

Name2016 7th International Conference on the Network of the Future, NOF 2016

Conference

Conference2016 7th International Conference on the Network of the Future, NOF 2016
Country/TerritoryBrazil
CityBuzios, RJ
Period16/11/1618/11/16

Keywords

  • Adaptive Network Coding Redundancy
  • MPEG-4/H.264 Scalable Video Coding (SVC)
  • Wireless Networks

Fingerprint

Dive into the research topics of 'Adaptive redundancy scheme for scalable video traffic using network coding'. Together they form a unique fingerprint.

Cite this