Characteristic time based routing in information centric networks

Abstract

Information centric networking (ICN) aims to transform today's Internet from a host-centric model to a content-centric one by caching content internally within the network at storage-enabled nodes. Recently, multiple routing and cache management strategies have been proposed [1]-[6] to improve the user-level performance (e.g., content-download latency) in ICN. In this paper, we propose a simple routing strategy that leverages the concept of characteristic time to improve content-download latency. Characteristic time for a content in a cache indicates the amount of time in future a recently accessed content is likely to remain in that cache. Our proposed algorithm (CTR) uses characteristic time information to forward requests to caches where the content is likely to be found. CTR augments native routing strategies (e.g., Dijkstra's algorithm), works with existing cache management policies and thus can be implemented in ICN prototypes with minimal effort. We perform exhaustive simulations using realistic Internet topologies (e.g., GEANT, WIDE, TISCALI, ROCKETFUEL [7]) and demonstrate that the CTR algorithm provides approximately 10-50% improvement in latency over state-of-the-art routing and caching management strategies for ICN for a wide range of simulation parameters.