Research On The Service Cache Methods In Information-Centric Networking

The traditional Internet communication paradigm with the center of host network address has severe problems in network scalability, security, manageability, controllability, and mobility, which cannot adapt to economic and social development needs of information networks. Thus, in recent years, Information-centric networking (ICN) has attracted widespread attention in the industry. ICN names information, by which information can cache in every node without changing name. Therefore, it is much easier for users to retrieve content, and it improves network efficiency and user experience. This paper mainly focuses on the research of caching mechanisms in ICN and studies the corresponding strategies, aims at improving the caching efficiency of the network. The main contributions of this paper are as following:1) To choose the proper naming granularity in ICN naming scheme, this paper analyzes the impact of content naming granularity on caching. Based on the naming scheme of content, this paper gives the definition of content naming granularity and analyses the strategy of content splitting and the advantage of distributed caching as well as the influence of naming granularity on the scalability of network. On this basis, this paper performs the simulations, which focus on the impact of content naming granularity on caching hit ratio, content transfer hops and delay with different parameters like topology, consumer request model and caching space. The result shows cache efficiency improves with the decrease of naming granularity, but the improvement is very small when naming granularity is fine. While fine naming granularity will bring network scalability problems. Based on the parameters, when naming granularity is between [10MB-100MB], the impact of scalability problem is limited but the cache efficiency is still good.2) To give better support to stream media contents delivery and transfer, this paper proposes a segment-popularity based caching approach for stream media contents in ICN. By analyzing the access pattern of consumers for large size stream media contents, this paper concludes that different segments of a single content have different popularity, further proposes the consumer request model for large size stream media contents. As segments of a single content cache in different cache nodes in ICN, this paper proposes a segment level-based implicit cooperative caching approach, which assigns a cache router a level based on its location and archives implicit cooperation with tags in interest and data packets. The simulation results show that the proposed caching approach has better cache performance than other approaches, which takes advantage of segment-level popularity. The proposed approach has higher caching hit ratio, lower request delay and content transfer hops, which will improve the user experience on stream media service in ICN.3) In recent years, many in-network caching approaches have been proposed in the literature. To the best of our knowledge, however, these approaches do not consider the failure of routers in the network. Accordingly, their performance could be significantly affected by node failures. To fill this gap, this paper presents a robust in-network caching approach for ICN. In RINA, the nodes in a network are divided into two groups:a backup node and some normal nodes. To provide robustness, this approach caches some most popular content chunks at both the given backup node and the normal nodes so that the failure of any node in the network does not significantly affect the caching performance such as caching hit ratio. We carry extensive simulations to compare this approach and some existing approaches. The results show that RINA outperforms them in robustness and has good cache hit ratio.