Research On Routing Mechanism Of Content Centric Networking

Current TCP/IP networks had many challenges in terms of scalability,manageability,mobility,security and so on.It was unable to meet the users' demand for the rapid distribution and acquisition of content.Researchers believed that an important change for the future networks is to transform from the "host-centric" to the "content-centric".They proposed a new type of network architecture which is Content Centric Networking(CCN)to satisfy the needs of future new applications.CCN changes the IP layer of the traditional TCP/IP model to the content layer.It completes positioning,routing,and transmission based on content names rather than IP addresses and provides content services for users.At the same time,CCN expands traditional nodes with only forwarding capabilities into CCN nodes with both storage and forwarding capabilities.So it improves network transmission capabilities.One of the key issues of CCN is how to design an efficient routing mechanism.The traditional CCN routing mechanism only considers how the interest packets find the content data on publisher servers,but doesn't consider the real-time network status and the impact of a large number of copy caches.Therefore,the average request content delay of traditional routing mechanisms is very large,and also distributed copy cache resources are wasted.This thesis proposes two improvements for the traditional CCN routing mechanism,and carries out experimental simulation and performance analysis.The main contents of the thesis as follows:(1)Edge Node Potential-Enhanced Routing(ENPER).The routing mechanism,on the one hand,designs the potential model of different nodes in network.By increasing the potential of the edge nodes,the interest packets are attracted to the cache nodes located at the edge of the network as much as possible.It reduces the delay of the average request content and the load of publisher servers.On the other hand,the number of packets of the same interests are counted at the edge node.Distinguishing contents of different popularity can effectively reduce the occupation of network bandwidth by the advertisement packets.Simulation results show that ENPER is reduced 43% in the average request content delay compared to the traditional CCN routing mechanism of Best-routing.(2)Software Defined Status Aware Routing in Content-Centric Networking(SD-STAR).The mechanism uses a controller to collect network status information such as request latency,node load,and hops to publishers.Then calculate the network comprehensive performance value and forwarding probability,and update the forwarding priority of relevant interfaces.The interest packet then selects an interface with the highest priority for forwarding.The simulation results show that compared with the traditional CCN routing mechanism,when the rate of interest packets sent by the requesters nine per second,the total number of interest packets in the network can be reduced by 72%,so the network overhead is reduced.The thesis proposes improved content-centric networking routing mechanisms from distributed and centralized routing architectures,and also form cache content awareness and network state awareness.From the simulation results,we can see that the design of the thesis,which are ENPER and SD-STAR,both improve the performance of CCN's routing,and reduce the request delay and the node load.They build a technical foundation for theoretical research and technical development of the future content-centric networking routing mechanisms.