How to Make Content-Centric Networks Interwork with CDN Networks Wei You, Bertrand Mathieu Orange Labs {wei.you, bertrand2.mathieu}@orange.com Gwendal Simon Telecom Bretagne gwendal.simon@telecom-bretagne.eu Abstract—The Internet architecture, based on end-to-end con- nections, had difficulties to efficiently deliver the always increas- ing number of contents. Content Delivery Networks (CDN) have been deployed to improve the delivery. Recent research works propose a new networking architecture, much more adapted to the current Internet usage (end-users just care about the contents they want and not about the endpoints that provide them). The Content-Centric Networking (CCN) aims at replacing the IP paradigm. CCN has been investigated by the research community for few years, and some demonstrators have proved its feasibility. However, less works have addressed the CDN use-case so far. In this paper, we show that the current CCN design, which does not allow negative reply, is not suitable for interconnection with the CDN service. We then propose the use of two new tables in the CCN nodes that are interconnected to the CDN servers, in order to detect possible misses in the CDN servers and subsequently forward the requests for missing contents toward the original servers instead of the CDN surrogates. The evaluations we have performed highlight that the integration of both new tables does not incur an increase of the memory requirement and thus a cost in the node. The proposed solution is thus a viable solution to make CCN network work in close cooperation with CDN networks, each one keeping its specific business models and functions: transport for CCN, storage for CDN. I. I NTRODUCTION The Internet success has changed the way we work, learn and play. Internet is widely used and leads to an increasing number of contents to be delivered by the network [1]. However, the Internet architecture, which has been designed in the 60s-70s and is based on end-to-end connections, faces a complexity to sustain the continuous traffic growth. Content Delivery Networks (CDN) were deployed in the last years to improve the content delivery, by replicating the most popular contents on some specific repositories, as close as possible to end-users. CDNs are now largely used and there is a clear business model for CDN providers and network providers: the CDN providers offer the storage function, while the network operators offer the transport function. In the last years, network scientists have investigated a new networking approach, the Content-Centric Networking (CCN) [2]. This paradigm is based on the fact that nowadays users care more about the contents that they want but no longer about the endpoints who provide. Demonstrators have proved the feasibility of CCN, such as VoCCN [3], video delivery over CCN [4]. However, not many works address on the cooperation between CCN and CDN issue. Indeed, some researchers argue that CCN nodes includes native caching functions and thus can be CDN nodes themselves. But this way of thinking means that the hybrid CCN/CDN node should be able to ensure high capacity of routing and forwarding functions, as well as efficient content storage and management. As illustrated by [5] which concludes that a fast CCN router cannot support a big volume cache space or at prohibitive price, we believe that for performance issues, the two functions should not be mixed within one equipment. Furthermore, each actor (network operators and CDN providers) would not like to loose their current business models. In this paper, we show the current CCN design is not suited to efficiently interwork with CDN repositories, because the native CCN design does not allow negative reply. In case of content miss in the CDN repository, the latter cannot inform the CCN node that it cannot serve this content. We then propose an evolution of the internal CCN node design, which is interconnected to the CDN repository. The evolution includes the use of two new tables: one to send requests toward the CDN repository (Repository Forwarding Table) and one to detect possible miss in the CDN server (Pending Repository Interest Table). Our evaluation highlights that the integration of both new tables does not imply more memory space than the current CCN node, and thus it does not incur a higher cost. We also study different implementing locations of our new node, namely cRouter, in a hierarchical topology. Since other researchers have different opinions regarding CCN/CDN interconnection, we present in a dedicated section, the others approaches and their advantages/drawbacks, compared to ours. The rest of the paper is organized as follows. We detail the motivations, the cRouter architecture and the Interest/Data processing algorithms in Section II. We give some performance evaluations of our cRouter system in III. Then we discuss other solutions in comparison to ours in Section IV before concluding this paper in Section V. II. CROUTER: A EXTENDED CCN NODE SUPPORTING CDN REPOSITORY This section presents our cRouter solution, enabling the interconnection the CCN node with the CDN repository. A. An overview of Content-Centric Network The work about CCN is motivated by an observation that the Internet users now care more about which contents or infor- mation they are interested in than about where the information actually exist. However today’s IP based architecture relies on a host-to-host conversation model. CCN aims at replacing this host-to-host design by a new client-to-content model.