论文部分内容阅读
随着流媒体应用程序在Internet上的广泛应用,连续媒体将超过源服务器上可用数据的50%。数字视频的高带宽和实时性要求,必将给互联网的负载带来巨大的变化;单播目前仍是Internet上的一种主要的流媒体播放方式,而基于包交换的Internet不是为实时、不间断的流媒体传输而设计的,访问流媒体的用户必须忍受高启动延迟(Start-Up Latency)和不可预知的播放质量。因此,要在Internet实现规模化的流媒体应用系统上,必须走出点到点、服务器到客户这种流媒体传输模式,流媒体缓存(Caching)就是这种解决方案的一个重要研究方向。
As streaming applications become more widely available on the Internet, contiguous media will exceed 50% of the available data on the origin server. The high bandwidth and real-time requirements of digital video will inevitably bring huge changes to the load on the Internet. Unicast is still a major streaming media playing method on the Internet. The Internet based on packet switching is not real-time, not Designed for intermittent streaming, users accessing streaming media must endure Start-Up Latency and unpredictable playback quality. Therefore, on the Internet to achieve large-scale streaming media applications, it is necessary to go from point to point, server to client streaming media transmission mode, streaming media caching (Caching) is an important research direction of this solution.