DocumentCode
3042746
Title
Probabilistic Packet Scheduling Scheme for Hybrid Pull-Push P2P Live Streaming Protocols
Author
Alghazawy, Bahaa Aldeen ; Fujita, Satoshi
Author_Institution
Grad. Sch. of Eng., Hiroshima Univ., Hiroshima, Japan
fYear
2011
fDate
Nov. 30 2011-Dec. 2 2011
Firstpage
248
Lastpage
251
Abstract
Hybrid pull-push P2P live streaming protocols are proposed to reduce the delay and the overhead of mesh based protocols and to improve the robustness of tree based protocols. In conventional hybrid protocols such as Cool streaming, the partition of a given stream into several sub streams is conducted in a static manner and each peer selects its parents without considering their upload availability. In this paper, we propose a new packet scheduling scheme for hybrid protocols based on the notion of randomization. Our scheme attaches a pseudo random number to each packet, and associates a sub range of random numbers to each parent and-child pair so that a packet attached a random number will be forwarded through a link associated with a sub range containing the random number. The size of each sub range is determined by considering the upload availability of each parent. By taking such a dynamic, probabilistic approach, we could improve the efficiency of resource utilization, as well as the stability, flexibility, and the robustness against peer churns.
Keywords
media streaming; peer-to-peer computing; probability; protocols; scheduling; cool streaming; hybrid protocols; hybrid pull-push P2P live streaming protocols; probabilistic packet scheduling; randomization; resource utilization; tree based protocols; Availability; Bandwidth; Delay; Peer to peer computing; Protocols; Robustness; Servers; Hybrid pull-push protocol; live streaming; peer-to-Peer system; scheduling;
fLanguage
English
Publisher
ieee
Conference_Titel
Networking and Computing (ICNC), 2011 Second International Conference on
Conference_Location
Osaka
Print_ISBN
978-1-4577-1796-3
Type
conf
DOI
10.1109/ICNC.2011.44
Filename
6131814
Link To Document