A new design for Information Centric Networks

Abstract

Information Centric Network (ICN) is a content-based information dissemination approach that improves the content delivery and latency. We introduce a modified architecture for ICN which can reduce the traffic by combining multiple messages and facilitate the data distribution in the network. By observing the similarities between index coding and ICN architectures, we propose to combine these two techniques to arrive at a new architecture that enhances data delivery in networks beyond original ICN scheme. By taking advantage of some concepts such as linear network coding, caching and index coding, we demonstrate that we can reduce the traffic and increase the capacity of ICN architecture by combining (network encoding) multiple messages requested by different nodes and sending them in one transmission. We demonstrate that each node will be able to extract its desired message from the combined encoded messages. To achieve this goal, we first define a modified version of index coding in order to apply index coding for both wired and wireless networks. Further, we introduce a hybrid caching scheme that includes both central and distributed caching to support two different goals. Our hybrid caching approach is a combination of conventional caching in ICN that caches the content in various network locations in order to make the content readily available to nodes and a new distributed caching scheme across nodes in the network to improve the performance of the entire system. The purpose of the second caching scheme is to allow central cache system to combine multiple contents in order to serve several client nodes simultaneously with a single transmission of encoded messages. The focus of this paper is to describe the new ICN architecture and demonstrate the advantages of the new architecture.