Low Inter-Reference Recency Set Replacement Policy in Named Data Networking

Abstract

Information-Centric Network (ICN) is a promising network architecture with content name-based routing and in-network caching as its main features. The characteristics of ICN are very suitable for future Internet with frequent information interaction. The performance of ICN is highly related to in-network caching, therefore, selecting a lightweight and efficient cache replacement policy is of vital importance in ICN. In this paper, we implement LIRS algorithm as a new cache replacement policy in Named Data Networking (NDN), one of the most mature ICN architectures. In ndnSIM, we simulate and compare the replacement policies of LIRS, LRU and FIFO based on different cache size, number of contents and Zipf parameters. Simulation results show that LIRS policy greatly improves the cache hit rate within O(1) time complexity and outperforms LRU policy and FIFO policy under different experiment settings.