Analysis Model of Probabilistic Caching Scheme for Content-Centric Networking

摘要

Even though a probabilistic caching scheme has high potential to be a caching scheme for Content-Centric Networking (CCN) due to its simplicity and effectiveness, there has been little work on analyzing its logical behavior in single-cache and multi-cache contexts. In this work, we propose an analytical model to evaluate the performance of the probabilistic caching scheme with Least-Recently Used (LRU) replacement policy. We first construct a Markov chain of a single cache based on Independent Reference Model (IRM) and extend it to the case of two-level cache hierarchy. We use the model to compute local cache hit rate, a parameter indicating the performance of a single cache. Moreover, our model reveals some insight into the cache diversity, an important factor contributing to global cache hit rate of a network of caches. Results from computer simulations based on ndnSIM validate the accuracy of our proposed model.