While the Internet has far exceeded expectations, it has also stretched initial assumptions, often creating tussles that challenge its underlying communication model. Users and applications operate in terms of content, making it increasingly limiting and difficult to conform to IP's requirement to communicate by discovering and specifying location. To carry the Internet into the future, a conceptually simple yet transformational architectural shift is required, from today's focus on where ? addresses and hosts ? to what ? the content that users and applications care about. This project investigates a potential new Internet architecture called Named Data Networking (NDN). NDN capitalizes on strengths ? and addresses weaknesses ? of the Internet's current host-based, point-to-point communication architecture in order to naturally accommodate emerging patterns of communication. By naming data instead of their location, NDN transforms data into a first-class entity. The current Internet secures the data container. NDN secures the contents, a design choice that decouples trust in data from trust in hosts, enabling several radically scalable communication mechanisms such as automatic caching to optimize bandwidth. The project studies the technical challenges that must be addressed to validate NDN as a future Internet architecture: routing scalability, fast forwarding, trust models, network security, content protection and privacy, and fundamental communication theory. The project uses end-to-end testbed deployments, simulation, and theoretical analysis to evaluate the proposed architecture, and is developing specifications and prototype implementations of NDN protocols and applications.
The goal of the NDN project is to redesign the Internet to offer better support for security, mobility and new applications. In the NDN Internet, every piece of data has a unique name and users retrieve data by the names. The data can be stored anywhere and served by any device that has the data, which allows much more efficient and scalable data distribution. For example, the network only needs to deliver a YouTube video once over each link, even to millions of users watching at different times. In addition, every piece of data is signed by its creator and can be verified by anyone. The creator can also encrypt the data so that only people with the right key can unlock it. In summary, security is a built-in feature of the NDN Internet, not an after-thought. Our specific research focus is on developing routing protocols that will help guide Interest packets to retrieve Data. Routing in NDN is different from traditional IP routing in two ways: (a) routing on names: destinations are name prefixes, not address prefixes; (b) multipath: routing protocol can provide multiple paths to each name prefix (if such paths exist). At the beginning of the project, because other NDN research areas (e.g., application and network management) need to use the NDN testbed for their prototyping and evaluation, we quickly developed a short-term routing solution "OSPFN" for the testbed by extending the widely used IP routing protocol OSPF to provide the above functionality. In the next few years, we developed a native NDN routing protocol called NLSR (Named-data Link-State Routing Protocol) http://named-data.net/doc/NLSR/0.1.0/), which has been deployed on the NDN testbed. NLSR is a named-based multipath routing protocol that supports two types of routing algorithms - link state and hyperbolic routing, as well as a hierarchical trust model for routing within a single administrative domain. We implemented hyperbolic routing in NLSR by disseminating hyperbolic coordinates in link state announcements. We have conducted extensive Emulab experiments to evaluate the feasibility of hyperbolic routing in NDN by comparing it with link-state routing under various conditions. Our results are archived at http://netwisdom.cs.memphis.edu/pvthome.html. Furthermore, we contributed to the following work in collaboration with University of Arizona and UCLA: (1) development of adaptive forwarding strategies: this work compared NDN, IP, and Path Splicing in face of pre x hijack, link failure, and congestion in simulations. The results show that NDN's forwarding strategy can effectively detect these problems and recover from them; (2) understanding of routing's role in NDN: this work compared routing algorithms (e.g., link state, distance vector) under IP and NDN in simulation, and showed that NDN routing can be simpler and more scalable because routing churns in an NDN network are handled locally by forwarding instead of globally by routing; and (3) routing scalability: we sketched out a solution similar to the map-n-encap approach in scaling IP routing - basically we map application names to routable names (typically belonging to ISPs, large companies and large content providers). This mechanism can also be used to support mobility and overlay construction. Finally, we participated in the development of NFD (NDN's forwarding daemon, http://named-data.net/doc/NFD/current/). More specifically, we designed and implemented the RIB manager, which registers name pre fixes received from applications and routes from routing protocols in NFD's FIB.
