Issue No. 04 - April (2009 vol. 20)
DOI Bookmark: http://doi.ieeecomputersociety.org/10.1109/TPDS.2008.102
Yuan He , Hong Kong Universtiy of Science and Technology, Hong Kong
Yunhao Liu , Hong Kong Universtiy of Science and Technology, Hong Kong
Most P2P Video-On-Demand (VOD) schemes mainly focus more on mending service architectures and optimizing overlays but do not carefully consider the user behavior and the benefit of prefetching strategies. As a result, they cannot better support VCR-oriented services in terms of substantive asynchronous clients, and free VCR controls for P2P VODs. In this paper, we propose VOVO, VCR-oriented VOD for large-scale P2P networks. By mining associations inside a video, the segments requested in VCR interactivities are accurately predicted based on the information collected through gossips. Together with a hybrid caching strategy, a collaborative prefetching scheme is proposed to optimize resource distribution among neighboring peers. We evaluate VOVO through extensive experiments. Results show that VOVO is scalable and effective, providing short startup latencies and good performance in VCR interactivities.
Protocols, Distributed Systems, Distributed applications
Y. Liu and Y. He, "VOVO: VCR-Oriented Video-on-Demand in Large-Scale Peer-to-Peer Networks," in IEEE Transactions on Parallel & Distributed Systems, vol. 20, no. , pp. 528-539, 2008.