基于社会属性的CDN-P2P系统的邻居节点选择算法

In the CDN-P2P hybrid system which based on social attributes, the nodes will select the neighbor nodes from the P2P subnet which organized by the surrogate servers, the algorithm of node selection will effect the overall throughput and the bandwidth ratio. However there are problems exist in the current file share systems, such as poor exchange rate or low bandwidth usage ratio. In this paper we develop a dynamic node selection algorithms which based on social attributes. In the algorithm, we select the neighbor nodes according to the upload and download capability, the similarity of the require contents. and in the process of running, we adjust the neighbors continually. finally we run simulations to show that, this algorithm can raise the hit ratio of the resources, and then save the cache resouces of the surroagte servers, lower the consumption of the network transmission.