Research On Caching Strategy In Content-Centric Networking

As the expansion of the cover of the terminal, the traffic ratio generated by content rise sharply, so the traditional host-centric network is becoming more and more difficult to deal with the requirements of the high-level content-centric applications. Under this background,content-centric networking emphasizes content was proposed. In the content-centric networking, replacing the IP packets in the traditional network with interest packets and data packets uniquely identified by the content names, that is, the routing process of the interest packets is the retrieval process of the content, and pay more attention to the content itself. CCN routers with in-network caching capabilities can directly respond to requests for interest packets, thus reducing network load and improving network performance to a certain extent. How efficient and flexible use of the cache so that more Interest packages can get a response in in-network caching is the problem to be solved caching strategy research.Based on the difference of the popularity of different content in reality and the centrality of different nodes in the topology, the idea of hierarchical difference caching is put forward. 'Hierarchical' means the CCN routing nodes in the network topology differentiated by the hierarchical window function through their Betweenness Centrality, and'difference caching' means the cache probability is different in different layers. The cache probability is subdivided into the layer cache probability and the content cache probability. The layer caching probability can be flexibly controlled by the hierarchical window function,which makes the cache probability different in different network nodes.The content caching probability can be flexibly controlled according to the popularity of the content, and it has uniformity in different nodes.Through the layer cache probability and the content cache probability can make the network cache is used more efficient and flexible.This paper proposes an HD-Prob (p) caching strategy based on the idea of hierarchically different caching. In order to solve the problem that the probability of caching in Prob (p) cache strategy lacks the difference in the nodes with different centralities, by introducing the cache probability, the probability of caching nodes with high Betweenness centrality is reduced, and the overall popularity of caching data packets is improved, and the performance is optimized.This paper proposes an HD-CL4M caching strategy based on hierarchical difference cache. In order to solve the problem that in CL4M cache strategy packets cached in the cache replaces frequently, by introducing content caching and layer caching probability, the frequency of packet replacement is reduced, and the overall popularity of cached data packets is improved, and the performance is optimized.Based on Icarus simulation platform, HD-Prob (p) and HD-CL4M caching strategies are simulated respectively. Simulation results show that the idea of hierarchical difference caching is optimized for both the Prob(p) cache strategy and CL4M cache strategy with improving the cache hit rate and reducing the average network delay.