Research On Caching Technology For Content-centric Networking

Current Internet architecture was designed in the 1960s and 1970s,aiming to make it possible to share hardware resources via Internet.With the reduction cost of production of hardware resources,people focus more on how to get the contents quickly,which puts higher requirements on the quality,reliability and security of Internet service.The weaknesses of traditional TCP/IP have been exposed.Therefore,academics have launched the exploration of the future network architecture.Content-centric Networking stands out from the representative architectures and becomes a hot research.Content-centric Networking is a reform of the TCP/IP,which changes the "end to end" model to a content-based communication and sets content as the first class.In CCN,users care more about the content itself rather than where it comes from.In-network caching meets this kind of demand of users.As long as the cache node stores the requested contents,the node may become a content provider.Caching technology not only reduces the traffic in network,but also reduces the transmission delay and the pressure of the content source.This paper presents the research of the caching technology in CCN,mainly to solve the problem of cache utilization in CCN to improve the performance of the entire caching system.This paper describes the current status of TCP/IP architecture and its problems.Several typical future network architectures have also been introduced.Chapter 2 introduces the basic idea,architecture and key technology of CCN.2.2 focuses on the characteristics of the cache in CCN and the points to be researched.Cache ubiquitous,transparency and volatile make it difficult to make full use of caches.In traditional CCN strategy,only caches on the delivery path can be used to serve the request,caches off the path cannot be used,which makes the cache system inefficient.To solve this problem,chapter 3 introduces a tracker-based cache utilization strategy,aiming to expand the visible scope of caches and improve cache utilization.Tracker-based cache utilization strategy divides the whole cache system into several sub-systems.Each of the sub-systems has a tracker server as the controller to collect the caching information.This strategy makes a trade-off between the scopes of object availability to request and the incurred overhead in order to improve the cache hit ratio and reduce the transmission delay.Chapter 4 introduces a distributed caching strategy to improve the caching system performance as well.Distributed caching strategy sets tracker for each file to collect the content chunk.When cache hit counts reaches the threshold,the cooperative caching mechanism will be triggered.Besides,this strategy divides the caches into two different states:collaboration state and local state.The caches which are in collaboration state are set a certain protection time to avoid being replaced in the guard time.Caches in the collaboration state can participate in collaborative mechanism.According to the analysis,distributed caching strategy has higher cache hit ratio,lower transmission delay and network load.