Cluster-based In-networking Caching for Content-Centric Networking

With the Internet architecture changing from host-centric communication model to content-centric model, Content Centric Networking (CCN) has emerged. One distinctive feature of CCN infrastructure is in-networking caching. As cache capacities of routers are relatively small compared with delivered data size, one challenge of in-networking caching is how to efficiently use the cache resources. In this paper, we proposed a cluster-based in-networking caching mechanism to improve the cache hit ratio and reduce caching redundancy for CCN. We designed the improved K-medoids cluster algorithm to cluster the whole network into k clusters and Virtual Distributed Hash Table (VDHT) to efficiently control and manage the resources stored in each cluster. We also proposed different policies for intra cluster routing and inter cluster routing to effectively forwarding requests. Compared with representative on-path caching schemes and hash scheme by simulation, we concluded that our clusterbased in-networking caching mechanism can improve the cache hit ratio and reduce link load of networks.