Improving Content-Centric networks performance with progressive, diversity-load driven caching

Abstract

Today, the main usage of the Internet has become increasingly about content generation, sharing, retrieval and massive distribution. Widely used P2P applications, UserGenerated Content (UGC) (e.g., Youtube, Flickr) and increasingly popular Social Networks make the Internet service model become more content-oriented than host-oriented as it has been this far. To match the Internet infrastructure to this service model, a new Internet architecture has been proposed under the generic name of Content-Centric networking (CCN, see also Named-data Networking or NDN). One of the main features of this architecture is its ability to cache packets in intermediate routers to take advantage of spatio-temporal locality in serving multiple requests for the same content, reducing thus considerably the transit traffic. In this paper, we propose a new in-network caching scheme for CCN and evaluate its performance by comparing it to that of the default proposed policy via simulation.