Presentation
Optimal incentive-based scheduling of layered video packets in P2P streaming
Optimal incentive-based scheduling of layered video packets in P2P streaming
(2010)
Abstract
In Peer-to-Peer (P2P) file sharing systems, such as BitTorrent, peers measure contributions of other peers and upload packets only to a selected subset of peers based on their past contributions. This built-in incentive scheme has alleviated free-riding problem for P2P file sharing. However, it cannot be easily adapted to P2P streaming of scalable videos, because it assumes that each packet is equally important for every peer. This assumption is not true: the importance of a packet to a playback peer depends on a number of factors, including which video layer the packet belongs to, how close it is to the playback deadline, and how many neighboring peers hold this packet. In this paper, we formulated peer selection and packet scheduling process as an integer programming problem. The objective is to maximize the aggregate gain of the packets, measured by their "importance" perceived by playback peers. We evaluated the proposed optimal scheduling algorithm by conducting the experiments in a global scale Internet testbed, PlanetLab, and demonstrated its excellent performance.
Keywords
- incentives,
- layered video coding,
- Peer-to-peer networks
Disciplines
Publication Date
August, 2010
Location
Changsha, China
DOI
10.1007/978-3-642-14932-0_48
Comments
August 18-21, 2010
Citation Information
Xiao Su. "Optimal incentive-based scheduling of layered video packets in P2P streaming" Optimal incentive-based scheduling of layered video packets in P2P streaming (2010) Available at: http://works.bepress.com/xiao_su/26/