A New Cache Management Scheme in Information Centric Networks

Abstract

As one of the main characteristics of ICN (Information Centric Networking), in-network caching had played a huge role in reducing network traffic. How to make reasonable use of the limited cache space in order to improve the cache hit ratio and reduce the network traffic had become the focus of the cache management in ICN. Considering users had different preferences in different data types, and the impact of data popularity on the data access rates, a new cache management scheme was proposed for ICN. In the proposed scheme, data were classified into different categories based on their semantic information, and the interest vector of nodes was defined according to the corresponding user preferences in different categories of data in order to address replica placement. Meantime, data popularity was introduced to solve the problem of cache replacement. Simulation results show that the proposed scheme was both feasible and effective in saving the cache space, improving the cache hit ratio and reducing network traffic load.