Supporting the Web with an information centric network that routes by name A. Detti ⇑ , M. Pomposini, N. Blefari-Melazzi, S. Salsano CNIT, Department of Electronic Engineering, University of Rome ‘‘Tor Vergata’’, Via del Politecnico 1, Rome, Italy article info Article history: Received 26 February 2012 Received in revised form 9 July 2012 Accepted 9 August 2012 Available online 20 August 2012 Keywords: Internet architecture Future internet Information-centric networking Routing Caching Scalability abstract Information Centric Networking (ICN) is a new paradigm in which the network layer pro- vides users with content, instead of providing communication channels between hosts, and is aware of the name (or identifiers) of the contents. A fundamental ICN operation is the routing of content requests towards a node that is able to provide the requested content. To meet this goal, different routing architectures have been proposed so far. In this paper, we consider a network that uses a routing-by-name architecture, i.e. con- tent requests are routed on the base of the content name by using a name-based routing table. We focus on the scenario of fetching Web contents, assuming to use ICN in place of traditional TCP/IP means. In this scenario we need to handle tens of billions of name- based routes, due to the high numbers of Web contents and to the limited aggregability of their names. Consequently, re-using the existing architecture of an IP router would result in two severe problems. First, the current Forwarding Information Base (FIB) tech- nology is unable to contain all name-based routes. Second, implementing a so large Rout- ing Information Base (RIB) requires a very costly hardware. In order to overcome these problems, we propose a routing-by-name architecture, named Lookup-and-Cache, where the FIB is used as a cache of routes, while the RIB is stored in a remote and centralized rout- ing engine. By analyzing real Internet traces, we prove the effectiveness of the proposed architecture, which we also show to be feasible with current technology. In fact, our ICN nodes require to have only a limited set of routes in their FIB, even when supporting a high number of traffic flows. We have implemented our proposed Lookup-and-Cache solution within the CCNx soft- ware framework and we used this implementation to assess system performance, such as download delay, lookup rate and fairness. The paper is completed with a discussion on how ICN can be used not only to fetch Web contents but also for other scenarios. Ó 2012 Elsevier B.V. All rights reserved. 1. Introduction Information Centric Networking (ICN) is a concept pro- posed some time ago under different names [1,2], which is attracting more and more interest, recently (see e.g. the papers [3–6], the workshop [36] and the projects [6–11]). ICN proposes a shift from the traditional host-to-host com- munication to a content-to-user paradigm, which focuses on the delivery of the desired content to the intended users. The basic functions of an ICN infrastructure are to: (i) address contents, adopting an addressing scheme based on names (identifiers), which do not include references to their location; (ii) route a user request, which includes a ‘‘destination’’ content-name, toward the ‘‘closest’’ copy of 1389-1286/$ - see front matter Ó 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.comnet.2012.08.006 ⇑ Corresponding author. Tel.: +39 06 72597285. E-mail addresses: andrea.detti@uniroma2.it (A. Detti), matteo.pompo sini@uniroma2.it (M. Pomposini), blefari@uniroma2.it (N. Blefari-Melazzi), stefano.salsano@uniroma2.it (S. Salsano). Computer Networks 56 (2012) 3705–3722 Contents lists available at SciVerse ScienceDirect Computer Networks journal homepage: www.elsevier.com/locate/comnet