NeTS: Small: Toward All Videos on Demand

视频点播(Video-on-Demand，VOD)是一项颇具吸引力的服务，它允许用户随时观看大型目录中的视频，从而在互联网上广受欢迎。现有视频点播服务的一个限制是，无论是使用传统设计还是P2P，它们都无法支持对所有高质量内容的高效流传输访问，而不考虑需求：不仅是数百万人想要观看的大片，而且还有与非常受欢迎的同行相比需求相对较小的小众内容：重播电视节目、非大片发行、体育赛事重播和演讲。对等(P2P)方法是可扩展内容分发的有效解决方案，而不会对集中式基础设施造成重大负担。在基于P2P的VoD系统中，用户从VOD服务器和对等点接收流媒体视频。观看视频的对等点之间相互协作的能力减少了服务基础设施的负担。该项目扩展了成功的P2P VOD系统的设计，以支持高效交付大型、潜在无限的高(电影)质量存储VOD库。无论视频的受欢迎程度如何，连接到网络的任何客户端都可以播放高质量的视频，而不会在其请求的几秒钟内中断。该研究采用了三个设计原则：利用P2P技术，实现利他的跨内容缓存和服务，使用回放点加权缓存策略，结果包括描述期望的缓存行为的模型，实现期望的缓存行为的分布式算法，以及功能原型。

Video-on-demand (VoD) is an attractive service that hasalready gained popularity in the Internet byallowing users to view a video from a large catalog atany time. A limitation of existing VoD services, bothusing traditional design as well as P2P is their inability to supportefficient streamed access to ALL content at high quality,regardless of demand: not just blockbusters that millions want tosee, but also niche content that has relatively small demand incomparison to their highly popular counterparts: re-runs of TV shows,non-blockbuster releases, replays of sporting events, and speeches.The peer-to-peer (P2P) approach is an effectivesolution for scalable content distribution without imposing asignificant burden on a centralized infrastructure. In a P2P-basedVoD system, users receive streamed videos from VoD servers as well asfrom the peers. The ability of peers that are viewing the videos tocollaborate with each other reduces the load on servinginfrastructure.This project extends the design of successful P2P VoD systems toenable support for efficient delivery of a large, potentiallylimitless library of high (movie) quality stored VoD. Regardless ofa video's popularity, a high-quality version will be playable by anyclient attached to the network without interruption within seconds ofits request. The research utilizes three design principles: leverageP2P technologies, implement altruistic cross-content caching andserving, and utilize playback-point weighted caching strategies.Results include models that describe desirable caching behavior,distributed algorithms to implement the desired caching behavior, anda functional prototype.