A Decentralized Pagerank Based Content Dissemination Model At The Edge Of Network

In the era of Internet of Things, cloud services are difficult to meet the real-time transmission requirements of users for the data generated in the edge of network especially for the Internet video services. Utilizing the devices at the edge of network, such as an intelligent router, to achieve nearby content services for users can effectively reduce backbone traffic and enhance service performance. This article proposes a decentralized PageRank-based content dissemination model at the edge of network, in which a suitable node selection algorithm is designed to distribute the content evenly in the network. Each node can quickly obtain data from neighbor nodes, thereby reducing the cloud load as well as the network bandwidth and improving the service response performance. The simulation shows that, compared with the other two dissemination algorithms, the content is distributed more even, which means every node has more opportunity to obtain the data from neighbors; and the service rejection rate can be decreased by an average of 5.2% in the case of high concurrent requests.