Streaming layered video over peer-to-peer (P2P) networks has been recognized as an efficient way to address the receiver heterogeneity problem. The distinctive characteristics of layer encoded video also introduce complexities to data scheduling. In this paper, we propose a new data scheduling algorithm for layered P2P video-on-demand (VoD) streaming networks. Our algorithm consists of two parts: 1) layer adaptation, where peers adaptively adjust the number of subscribed layers to ensure a continuous playback of the highest possible video quality; and 2) piece selection, in which a peer selects a missing data piece to request based on its utility. The piece utility is calculated based on the playback constraint and the layer dependency of a piece. Through extensive packet-level simulations we show that our proposed data scheduling algorithm can effectively enhance the video playback quality.
展开▼
