Least Recently Frequently Used Replacement Policy Named Data Networking Approach

Abstract

Named Data Networking (NDN) is a new paradigm of a network by using Content-Centric Network (CCN) theory. NDN provides better network performance since the caching policy of the NDN router saving the content requested by the user and make it accessible by another user. The current popular replacing policy that adopted from TCP/IP network is Least Recently Used (LRU) and Least Frequency Used (LFU). However, in a low condition of cache size, both algorithms not powerful yet. In this paper, we evaluate replacement policy in NDN using Least Recently/Frequency Used (LRFU) compared with LRU algorithm. Moreover, our approach based on C++ programming language to simulate the LRFU caching policy in Named Data Networking. Overall results shows that LRFU replacement policy in NDN provides higher hit rate and lower miss rate. Based on 20000 generated interest, LRFU replacement policy achieve 3578 average hit which is 18.398% greater than LRU with average hit 3022. Based on miss rate LRFU replacement policy achieve 16391 miss rate which is 3.28% lower than LRU with 16948 miss rate.