云计算与对等计算混合环境下流媒体分发关键技术研究

Due to the limitation of the dynamic and transmission control, the streaming distribution technology based on Peer-to-Peer (P2P) is difficult to steadily ensure system transmission QoS (Quality of Service). Since cloud computing has the characteristics of high storage space and bandwidth access, which can compensate the defect of P2P mode, the combination of these two patterns for streaming distribution is the major trend and opening research question in this research area. This project systematically study and optimize the streaming distribution key technologies in the combined cloud computing and P2P environments from the following three aspects: organizational architecture design, transmission performance optimization and resource deployment optimization. Through combining the characteristics of cloud computing and P2P, the project team provides a hierarchy mixture organizational architecture, builds and optimizes a tree-mesh hybrid transmission topology for the streaming system based on the cloud computing platform and P2P network; through formulating the optimization model, the project team proposes the optimal data transmission scheduling based on network coding and designs the cloud computing platform’s data migration mechanism, so that the transmission QoS of the system in hybrid environments could be effectively improved; based on the system demand for cloud computing resources, the project team establishes and solve the optimal problem of cloud computing’s resource deployment, in order to reach the optimal balance between the transmission QoS and the system operation cost. The research results of this project hopefully break through the bottlenecks in the two aspects of transmission performance and system deployment for the streaming distribution technologies in the combined cloud computing and P2P environments, realizing the large-scale deployment and application of this hybrid pattern system in actual network environments.

基于对等计算(P2P)的流媒体分发技术由于动态性和传输控制方面的劣势，难以稳定地保证传输服务质量。云计算提供稳定和大量的存储空间和带宽访问能力，可弥补P2P模式的缺陷，两者结合分发流媒体是当前该领域的主要趋势和开放研究问题。本项目针对云计算和P2P环境下流媒体分发关键技术，从组织架构设计、传输性能优化和资源优化部署三个方面，进行系统性研究。项目组结合云计算和P2P模式的各自特征，提出分层混合组织架构，建立和优化“树网”结合的传输拓扑；通过构建最优化模型，提出基于网络编码的最优传输调度算法，设计云计算数据迁移机制，提高云计算资源利用率和系统传输服务质量；根据对云计算资源的动态需求，建立和解决云计算资源部署最优化问题，达到系统传输服务质量和经济开销的最佳平衡。本项目研究成果有望突破云计算和P2P混合环境下流媒体分发技术所面临的传输性能和运营部署的双重瓶颈，实现系统的大规模实际部署和应用。

基于对等计算(P2P)的流媒体分发技术由于动态性和传输控制方面的劣势，难以稳定地保证传输服务质量。云计算提供稳定和大量的存储空间和带宽访问能力，可弥补P2P模式的缺陷，两者结合分发流媒体是当前该领域的主要趋势和开放研究问题。本项目针对云计算和P2P环境下流媒体分发关键技术，从组织架构设计、传输性能优化和资源优化部署三个方面，进行系统性研究。项目组结合云计算和P2P模式的各自特征，提出分层混合组织架构，建立和优化“树网”结合的传输拓扑；通过数据驱动的研究方法，提出了面向移动端视频的云辅助和P2P辅助流媒体分发传输技术，并对于云平台资源的更新和存储策略进行了有效设计；根据对云计算资源的动态需求，提出了针对云辅助和P2P辅助流下的基于多目标优化的云资源租用策略，达到系统传输服务质量和经济开销的最佳平衡。本项目研究成果对现有云计算和P2P混合环境下流媒体分发技术所面临的传输性能和运营部署的瓶颈有所突破，有利于增强该类型系统的实际部署和应用。