视频网格中流媒体业务的流量模型

Streaming Media Service Traffic Model in Video Grid

针对网格环境下的多服务器对单个用户提供服务的流媒体服务模式,提出了一种多层次的网格环境下的通用流媒体业务的流量模型——复合AR模型(MAR).根据媒体网格中流媒体的特点,将媒体流分为媒体流段和媒体流片断,媒体流片断是近似平稳过程,采用AR模型建模,媒体流段中的媒体片断之间符合半马尔可夫过程,媒体流中的媒体流段符合泊松过程.该模型为网格环境中提供更加有效、高质量的VoD服务打下了基础.

Proposes a multi-layer general streaming media service traffic model（multiple AR model,MAR） in video grid for the service pattern of multi-server serving single client.The media streaming is layered into media streaming segment and media streaming sub-segment.Media streaming segment as a homogenous process is modeled by AR,sub-segments in media streaming segment are described by semi-Markov stochastic process,and segments in media streaming accord with Poisson distribution.The model sets up a basis for supply more effective and high quality VoD service in video grid.