分层多描述编码在P2P流媒体点播系统中的研究与应用-计算机应用技术专业论文.docx

华中科技大学硕士学位论文 华 中 科 技 大 学 硕 士 学 位 论 文 II II Abstract The emergence of the media streaming makes the video be transferred over network continuously in real-time mode possible. Users need not have to wait until the whole media is downloaded，which meets users desire of real-time mutual communication. At Present, most media streaming servers are based on C/S model which limits their extensibility and reliability. However, the media streaming technology based on P2P(Peer-to-Peer) can overcome these limitations which makes the transportation and application with mass scale media streaming possible. Though there are already some successful P2P real-time broadcasting systems such as PPLive, PPStream, a successful P2P VoD system has not appeared yet. On the base of research on P2P streaming media VoD System (PPVOD), the author studied transportation of P2P network and coding of streaming media, think over layered description coding, and do his best to improve PPVOD system QoS. Proposed a sub-team partition mechanism and layered strategy based on performance value and introduced in detail the user model of PPVOD system in the paper. Layered strategy based on performance value can satisfy as many as the heterogeneous receivers, and users can choose to receive the different descriptions of video data in according to the state of network and the performance of machine, and different peer can enjoy the different QoS video. Analyzed layered coding, multiply description coding and layered multiply description coding, providing the approach a layered multiply description coding base on space domain splitting, and provided a specific codec processes and related algorithms. Given a simulation experiment to testify the approach a layered multiply description coding base on space domain splitting. The approach is proved very effective in reducing transmission of video data and improving PPVOD system QoS. Key words：Layered multiple description coding, Peer-to-Peer network, Streaming media