Media Streaming Caching Algorithm Based on Actual User Behavior Analysis

XIA Yan,WANG Song,AN Ran,XIE Tie-bing
DOI: https://doi.org/10.3969/j.issn.1000-1220.2012.08.013
2012-01-01
Abstract:Media Streaming Caching is an important method to alleviate the network bottleneck and the disk I/O bottleneck.Users show very complicated behavior when they are visiting the media streaming system.These behaviors are very critical to design effective caching algorithm.The current caching algorithms are based on the hypothesis that the users obey Poisson Distribution or Zipf Distribution.The hypothesis is usually not consistent with the actual instance.In this article,we design a caching algorithm of media streaming based on the users′ actual click behavior and action.We use actual data to experiment the algorithm,the result shows that the algorithm is effective.
What problem does this paper attempt to address?