A content-based caching algorithm for streaming media cache servers in CDN

Abstract

High-quality streaming is getting more popular as it is getting more attention of Internet users. This was possible because there were enough high-speed network infrastructure and Contents Delivery Network (CDN) service. So far, CDN has supported a streaming service by using streaming media cache server that is based on the web cache server as its platform. The static caching is the most popular to the web cache server dealing with the static content. However, the media object of the streaming cache is different from that of the web cache in terms of the size and effective duration. Streaming service requires a large storage space and demands more amount of the traffic. In particular, the traffic of the streaming service is more severely variable due to the faster and more frequent interaction between the client and the server. For these streaming services, CDN uses dynamic caching that can save the cache space and can reduce the response time on the demand of the user. However, the dynamic caching causes the cost of heavy CPU burden. In this paper, we propose a new caching algorithm based on the dynamic caching for streaming media cache servers. The proposed algorithm is a content-based approach to cache an individual content to a server according to the popularity of the content. That is, the contents of high popularity only can be distributed over the multiple cache servers. The experimental results show that the proposed algorithm performs better than conventional approaches in terms of the cache hit rate and the amount of the traffic across the network.