Design And Implementation Of Network Cache System For Video Applications

With the popularization of 4G technology and the rapid development of mobile devices such as smart phones and tablet PCs,mobile Internet-related Internet services are gradually becoming more dominant in the Internet.The data traffic generated by them also grows at an alarming rate,bringing tremendous pressure to the Internet Infrastructure.In the era of mobile Internet,how to speed up the distribution of web contents,increase the bandwidth utilization rate of the backbone network,and reduce the network delay of mobile users so as to enhance the user experience has become a problem that researchers are concerned about.In this paper,we designed and implemented a network caching demo system for mobile video applications,which achieves the effect of transparent caching by intercepting and analyzing user requests and using a flexible multi-level caching network to decentralize server-side resources.The system enables users to obtain content closer to the edge of the network,thus effectively reducing network latency,saving bandwidth resources,relieving service pressure and achieving efficient network content distribution.The network caching demo system in this paper can be divided into a highly flexible,scalable and easy-to-deploy caching system composed of network nodes with different functions such as caching nodes and index nodes,as well as a demo system using a global monitoring page as the display interface to show all the communication process between network nodes in real time.The system is implemented based on Linux/Node.js,with the user interception part based on iptables TCP redirection.In this paper,we use the system to test the automatic video playing of the two network architectures of ubiquitous distribution network and MCDN in the project,thus verify the effectiveness of the system.In addition,based on this network caching demo system,this paper extends the module for simulating user requests and control caching strategies to implement the experimental functions related to caching strategies.In this paper,the modified system is used to compare different caching algorithms under single node and experiments on different nodes with different network topologies,and the results are analyzed.