Research On Caching Strategy For Vod In Named Data Network

Since the 21 st century,with the rapid development of Internet technology and the maturity of multimedia technology,online video on demand is popular.However,with the development of the network,especially in the data traffic,the traditional video on demand has a problem that the video picture is not clear,the frame skipping,the continuity is poor,and the user experience is not good,except for the low network bandwidth and complicated topology.In addition to the video data unpopular and other reasons,the key to these problems is that the TCP/IP-based communication mode is a waste of serious resources on the Internet,that is,a large number of repeated(redundant)transmissions occur on the Internet..In response to the problems of IP networks responding to the increasing network traffic,the Named Data Network(NDN)came into being.The emergence of the NDN project provided a possibility to replace the TCP/IP protocol.The project established a new content-centric network architecture.In order to reduce congestion and improve content reuse,the NDN network can cache data content on each routing node to maximize bandwidth usage and achieve fast and reliable content delivery.In the NDN network,each routing node has an independent buffer space.When the data packet passes through the node,the node will cache the data locally,which greatly reduces the repeated traffic in the network and speeds up the response to the user request.However,the cache also puts pressure on the routers in the NDN network.Therefore,the cache policy of the NDN network must be well designed to better achieve the cache target.This paper aims at optimizing the NDN cache strategy for the characteristics of on-demand applications with large traffic in the network.It is hoped that the practicality of the NDN network will be promoted.This paper mainly studies from two aspects.One is the cache decision,that is,which router node the data decision returned from the server is placed on,which can make the placement effect better;the second is cache replacement,that is,when the router space usage exceeds a certain upper limit,Which secondary information needs to be deleted,leaving more room for useful information.This paper proposes [to summarize the two algorithms you proposed] hot data asymptotic user algorithm and content lifetime algorithm,the former as a cache decision algorithm,the idea is to complete the data content of the number of requests,that is,the hot content can follow the heat The increase will be infinitely close to the host user;the latter as a cache replacement algorithm,the idea is that when a routing node storage space is full,the priority is to replace the data content with the lowest content lifetime.In this paper,the simulation experiment of caching strategy is carried out onndn SIM.The cache decision strategy and cache replacement strategy proposed in this paper are compared with the decision strategy and replacement strategy adopted by NDN.It is found that the caching strategy proposed in this paper has multiple performances.The indicators have certain advantages,thus achieving the purpose of improving content reuse and reducing server pressure.In the end,the video on demand system interface is designed.An effective decision control layer is designed between the application layer and the NDN network layer.The interface between the application layer and the NDN network layer is used to complete the adaptation of the video on demand in the NDN network.