DocumentCode :
2076836
Title :
Adaptive Mechanism for P2P Video Streaming Using SVC and MDC
Author :
de Asis Lopez Fuentes, Francisco
Author_Institution :
Dept. de Tecnol. de la Informacion, Univ. Autonoma Metropolitana Cuajimalpa (UAM-C), Mexico City, Mexico
fYear :
2010
fDate :
15-18 Feb. 2010
Firstpage :
457
Lastpage :
462
Abstract :
In this paper we propose and evaluate a combined SVC-MDC (Scalable Video Coding & Multiple Description Video Coding) video coding scheme for P2P video multicast. The proposed scheme is based on a full cooperation established between the peer sites, which contribute their upload capacity during video distribution. The source site splits the video content into many small blocks and assigns each block to a single peer for redistribution. Our solution is implemented in a fully meshed P2P network in which peers are connected to each other via UDP links. The video content is encoded using SVC. We present a flow control mechanism that allows us to dynamically optimize the overall throughput and to automatic adjust the video quality for each peer. Thus, peers with different upload capacity receive different video quality. We also combine SVC with MDC to alleviate the packet loss problem. We have implemented and tested this approach in the PlanetLab infrastructure. The obtained results show that our solution achieves good performance and remarkable video quality in the presence of packet loss.
Keywords :
peer-to-peer computing; telecommunication congestion control; transport protocols; video coding; video streaming; P2P video streaming; PlanetLab infrastructure; UDP links; flow control mechanism; multiple description video coding; packet loss problem; scalable video coding; Bandwidth; Competitive intelligence; Delay; Peer to peer computing; Static VAr compensators; Streaming media; Throughput; Video coding; Video sharing; Web and internet services; multiple description video coding; peer-to-peer networks; scalable video coding; video streaming;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Complex, Intelligent and Software Intensive Systems (CISIS), 2010 International Conference on
Conference_Location :
Krakow
Print_ISBN :
978-1-4244-5917-9
Type :
conf
DOI :
10.1109/CISIS.2010.111
Filename :
5447463
Link To Document :
بازگشت