CPCache: Cooperative Popularity based Caching for Named Data Networks
Neminath Hubballi, Pankaj Kumar Chaudhary · 2024
Information Centric Networks improve content delivery performance by caching popular content in the network routers. However, estimating popularity comes with cost of every router exchanging content requests with every other node which leads to a significant overhead. In this paper, we describe CPCache which brings cooperation among a group of routers to estimate popularity and make caching decisions. It identifies a set of designated routers in the network to coordinate the popularity estimation and assist edge routers in making caching decisions. In CPCache, every edge router associates itself with a designated router from which it seeks assistance to (knowing whether a new content is popular or not) make caching decisions. We describe methods for selecting designated routers, edge router association with a designated router, content popularity estimation mechanism and also performing caching decisions by routers. We evaluate and benchmark the performance of CPCache with five other popular caching methods to conclude it offers better performance compared to these methods using simulations done with a popular discrete event simulator.