Video Replica Placement Strategy for Storage Cloud-based Cdn

The online video service need the support of CDN(Content Delivery Networks). Compared with traditional CDNs, it can save a lot of cost by using cloud-based storage nodes to deliver the video content. To guarantee end users’ QoS, CDN should pre-deploy the content files of online video service to the edge nodes which are close to the users. Existed researches have shown that the cost of building CDN by cloud storage nodes is much less than that of using traditional CDNs. The existed off-line replica placement algorithm named GS(Greedy Site) can meet the QoS requirement with relatively small cost when the information of users’ requests is provided. However GS will result in bad load balance and it need the information of users’ requests. In this paper, two classes of offline algorithms are proposed. One named GUCP(Greedy User Core Preallocation) effectively solved the load imbalanced problem caused by GS ,and the other one named PBP(Popularity Based Placement) which is based on the popularity of content effectively placed replicas while there is no users’ requests information. Numerical experiments have demonstrated the effectiveness of the algorithms above.