基于内容流行度的网络内部缓存智能分布方法

Intelligent Distribution of Cache in Network Based on Content Popularity

网络内部缓存的有效分布能够解决大数据量访问引发的网络拥堵问题。但传统缓存分布方法缺乏智能性,存在网络负载均衡度较低、内容冗余率较高,网络资源利用率较低的问题。为此提出一种基于内容流行度的网络内部缓存智能分布方法。根据网络内部缓存智能分布总体流程设计,分析网络缓存空间节点的布局,避免网络拥堵。通过节点的度和介数来计算节点的中心度,选取中心度值最大的节点,采用KNN算法对其内容流行度进行预测,对所得节点的内容流行度进行排名,完成网络内部缓存的智能分布。采用对比测试方式进行仿真实验,验证得出所提方法的内容冗余率为8.7%,网络资源利用率为92.6%,且网络负载均衡度也有所提高,具有一定的智能性。

The effective distribution of network internal cache can solve the network congestion caused by large data access.However,the traditional cache distribution method lacks intelligence,with the network load balancing degree being low,the content redundancy rate being high,and the network resource utilization rate being low.An intelligent distribution method of network internal cache based on content popularity was proposed in this paper.According to the overall process design of intelligent distribution of network cache,the layout of network cache space nodes was analyzed to avoid network congestion.The degree and intermediate of nodes were used to calculate the degree of centrality of nodes.The nodes with the largest value of centrality were selected,and KNN algorithm was used to predict the content popularity of the nodes and rank the content popularity of the obtained nodes for the purpose of completing the intelligent distribution of network internal cache.The comparison test method was used for simulation experiments.The results show that the content redundancy rate of the proposed method is 8.7%,that the utilization rate of network resources is 96.2%,and that the network load balance is also improved,with a measure of intelligence.