论文部分内容阅读
随机搜索是用户在享受视频点播服务时常见的行为,它使得播放进度发生突然变化,同时要求系统做出及时的响应.为了缩短数据访问的响应时延,通常采用预取机制提前获取部分数据用以满足将来的需要,然而如何利用有限的存储空间预取尽可能多的有用数据是其中的关键问题.本文根据视频点播用户随机搜索操作的行为特征模型和媒体内容不同部分受欢迎程度的不同,提出一种范围受限、流行度感知的P2P视频点播系统数据预取机制RC-PAP.仿真实验结果表明,相比于现有的数据预取方法,RC-PAP可以显著提高用户随机搜索操作时的响应速度,并减轻内容源服务器的访问压力.
Random search is a common behavior when users enjoy video-on-demand services, which makes the progress of broadcast changes suddenly, and requires the system to make timely response.In order to shorten the response time of data access, pre-fetching mechanism is usually used to obtain partial data However, how to prefetch as much useful data as possible with limited storage space is one of the key issues.In this paper, according to the difference of the popularity of different parts of the media content and the behavioral feature model of the video-on-demand user random search, This paper proposes a data-prefetch mechanism RC-PAP for P2P video-on-demand system with limited range and popularity.The simulation results show that compared with the existing data prefetching method, RC-PAP can significantly improve the user random search operation Response speed and reduce the pressure on the content source server.