电信科学 ›› 2015, Vol. 31 ›› Issue (1): 146-151.doi: 10.11959/j.issn.1000-0801.2015034

• 运营技术广角 • 上一篇    下一篇

基于流行度分类的流媒体分段策略

邓字珊1,陈戈2,张军1,梁洁2   

  1. 1 华南理工大学电子与信息学院 广州 510640
    2 中国电信股份有限公司广东研究院 广州 510630
  • 出版日期:2015-01-15 发布日期:2017-02-21
  • 基金资助:
    国家自然科学基金资助项目

Segmentation Strategy of Streaming Media Based on Popularity Classifying

Yushan Deng1,Ge Chen2,Jun Zhang1,Jie Liang2   

  1. 1 School of Electronic and Information Engineering, South China University of Technology, Guangzhou 510640, China
    2 Guangdong Research Institute of China Telecom Co., Ltd., Guangzhou 510630, China
  • Online:2015-01-15 Published:2017-02-21
  • Supported by:
    The National Natural Science Foundation of China

在线阅读

0

PDF下载

299

被引次数

1

摘要：

对流媒体内部、外部流行度的特点以及流媒体外部流行度与缓存替换次数的关系进行了分析，并在此基础上，针对流媒体的缓存替换通常集中在某一小段流行度范围的特点，提出了一种基于外部流行度分类的流媒体分段方法，令替换频繁的视频采用较小的分段，而其他替换不频繁的视频采用较大的分段，从而提高缓存空间的利用率。实验结果表明，与均匀分段相比，提出的流媒体分段方法可以用更少的分段数目达到相同的命中率。

关键词: 流行度, 分段, 流媒体

Abstract:

The characteristics of external and internal popularity of streaming media and the relationship of external popularity and the number of cache replacement were analyzed. Respecting the characteristics that cache replacements are focus on a small rang of popularity, a segmentation method based on popularity classifying of streaming media, which uses smaller segments on videos replaced frequently and bigger segments on the other videos, was proposed. The experimental result shows that the segmentation method can effectively reduce the number of segments compared with the uniform segmentation method under the same conditions of the byte hit ratio.

Key words: popularity, segmentation, streaming media

No Suggested Reading articles found!