A hierarchical cluster-based caching for named data networking

The in-network caching strategy in Named Data Networking (NDN) will reduce the burden of original content server, distribute the inter-AS traffic and decrease the response time. Nevertheless, there are many shortcomings to cache massive redundant copies, for instance, wasting of cache space, much replacement overhead and reduction of cache hit. With the purpose of overcoming these shortcomings and improving caching efficiency, we propose a Hierarchical Cluster-based Caching (HCC) strategy which has a two layers hierarchical architecture. The routers independently execute their respective caching decision in different layers. Furthermore, the routers in Core Layer do not have the caching function. However, the important routers in Edge clusters can cache contents according to betweenness centrality of each router, content popularity of required Data and the corresponding probability matrix which are calculated by clusterhead. We evaluate our strategy in ndnSIM, and the simulation results show that the performance of HCC is greatly outstanding.