Content-hierarchical intra-domain cooperative caching for information-centric networks

Information-centric networking (ICN) is a clean-slate network architecture to achieve efficient content transmissions. In-network caching is an important issue in ICN, since the routers have a limited caching capacity. Several approaches have been proposed to reduce the cache redundancy, and increase the cache hit ratio and the user delay. Most approaches design algorithms for single routers, and there are still redundant caches in an AS. In this paper, we propose the content-hierarchical Intra-Domain Cooperative Caching (IDCC) scheme to solve the problem, based on the observation that the time of access contents within the local AS is much shorter than that of access contents outside. We use probabilistic caching based on inlet traffic speed, and avoid unnecessary replicas by intra-domain cooperation based on content advertisements. We use hierarchical caching based on content popularity to improve the performance. Our scheme is different from other intra-domain cooperation schemes which remove redundant replicas after contents have been cached. We evaluate our scheme with simulations. The results show that IDCC can improve the hit ratio by 1.7 times and 1.1 times compared to two common existing schemes, respectively.