Research On Priority Service And Collaborative Caching Mechanisms In Information Centric Networking

As the proportion of multimedia services such as video in total traffic gradually increases,users' demands for networks gradually shift from end-to-end resource sharing to content distribution and acquisition.However,the current TCP/IP architecture is based on the the end-to-end communication model,which is obviously difficult to cope with the changes in users'needs.With the improvement of the router's cache capacity and computing power,Information Centric Networking(ICN)architecture has been proposed.In ICN,the separation of names and addresses is achieved by naming data at the network layer,routing based on content names in the network,and supporting the deployment of caches in core network nodes,so the efficiency of content distribution has been significantly improved.Therefore,ICN architecture is widely concerned.However,many problems have not yet been well solved,such as efficient utiliza-tion of the cache.Whether the cache can be effectively used depends on the design of the cache strategy.Therefore,research on the cache strategy has become one of the im-portant and difficult issues in the research work of ICN.Current caching strategies have not solved the following two issues:Firstly,existing caching strategies cannot provide priority-based service.The design of the cache strategy is considered from the perspec-tive of maximizing the benefits of Internet Service Providers(ISP).Because its purpose is to maximize the overall performance of the network system,there is no different treat-ment for different content providers(CP)and they all provide the same non-differential cache service.However,there is no corresponding mechanism to ensure the demand that some CPs want to pay ISP for better service.Secondly,the performance of the collaboration-based cache strategy is still not efficient enough and there is room for further improvement.Because users' content preferences are generally consistent and popular content will be repeatedly cached,which will bring greater cache redundancy and lower utilization of cache space.A lot of collaboration-based cache schemes have been proposed to solve this problem.However,the proposed scheme only achieves cache content sharing and do not realize the collaborative use of cache space,which is relatively coarse-grained.In addition,schemes do not consider the robustness of the system.Therefore,this paper deeply studied above two issues and separately pro-posed a cache strategy to provide priority-based service for CPs and a more efficient collaboration-based cache strategy.The main works and contributions of this article are as follows:(1)Aiming at the problem that existing cache strategies cannot provide priority-based service,this paper proposes a priority-based service scheme based on exclusive cache lease.It is different from the purpose and method of providing differentiated ser-vice for different types of contents in traditional networks.The core idea of the scheme is to comprehensively consider the two factors of node coverage and request delay to quantify the importance of nodes,and select the appropriate nodes in the network ac-cording to this value to complete the lease of the dedicated cache.For exclusive cache,a two-level popularity threshold cache strategy is proposed,which is used to store the global sub-popular content of the paying CP.It can ensure that the cached content is not replaced by other CPs' high-popular contents.Experimental results show that the scheme reduces the average delay of users requesting the paying CP,so the scheme provides priority-based service for the paying CP.(2)Aiming at the problem that current collaboration-based cache schemes only achieve cache content sharing but not achieve the collaborative utilization of cache space and the poor robustness,this paper designs a fine-grained and robust collaboration-based cache strategy,which makes full use of cache resources and further improves the cache performance.In this scheme,in order to consider the influence of the content re-placed by neighbor nodes and the cache copy of the same content on local nodes,an im-proved local replacement strategy is designed(which defines the cooperative value and duplicate value).In addition,in order to utilize the cache space cooperatively,a cache strategy based on neighbor collaboration is designed.These two strategies together con-stitute the neighborhood collaboration-based cache scheme in this paper.Experimental results show that compared with the traditional cooperative scheme,the average delay of users is further reduced and the cache hit ratio is further improved.In the case of node failure,the proposed scheme can guarantee the robustness of the system.