网络视频是互联网主流应用之一，本项目拟结合对等网络技术、CDN(Content delivery Network)与HTTP自适应流媒体技术，从内容分发角度研究构建低成本、高服务质量、支持多终端的流媒体视频点播云平台。首先，研究云端视频传输策略，以最小化云端视频传输费用为目标，提出基于修剪Dijkstra算法的费用最小路径算法。其次，研究混合CDN-P2P体系结构，构建跨域有序协作的CDN与P2P集成方式，构建基于改进幂律环状的P2P环状层次覆盖网。然后，以最大化视频播放质量和最大化网络吞吐量为目标，研究多目标优化的分布式客户端数据调度算法。最后，以最小化服务器带宽成本为目标，分别设计视频片段级别和视频级别的分布式客户端缓存管理策略。本项目的研究可为今后研发和部署低成本、支持多终端和HTTP自适应流媒体的视频点播云平台提供理论参考和技术借鉴。

Internet video has become a main application on Internet. This project is to explore how to construct a VoD (Video-on-Demand) Cloud with low-cost, high QoS and multi-terminal support from content delivery view, which can integrate P2P (Peer-to-Peer), CDN(Content delivery Network) and HAS (HTTP adaptive streaming). First, this project is to study the video delivery policy on the cloud side, aiming at minimizing the video delivery cost on the cloud side. To this end, a minimum-cost path algorithm based on an adjusted Dijkstra algorithm is proposed. Second，this project is to study hybrid CDN-P2P architecture，which can integrate CDN and P2P overlay with inter-domain collaborating on order. In the hybrid CDN-P2P architecture, P2P ring layered overlays based on an improved power law ring structure are proposed. Third, this project is to design a multi-objected optimal and distributed data scheduling strategy for the client nodes to achieve the two objectives: maximizing the streaming playback quality and maximizing the network throughput. Last, this project is to design a distributed cache management strategy for the client nodes from video segment level and video level, respectively, to minimize the server bandwidth cost. The research results of this project can provide theoretical reference and technical reference for developing and deploying a low cost HTTP adaptive streaming VoD cloud with multi-terminal support.