Named Data Networking

Named Data Networking
Name	Named Data Networking
Developer	UC Berkeley, UCLA, USC, PARC
Released	2010s

Named Data Networking is an architectural approach to future packet-switched networks that centers communication on retrieval of named data rather than on host addresses. It rethinks principles used in Internet Protocol, Transmission Control Protocol, User Datagram Protocol, and traditional OSI model stacks by emphasizing content-centric interactions similar to paradigms explored at PARC, USC Information Sciences Institute, and University of California, Berkeley. The project draws participation from researchers affiliated with institutions such as University of California, Los Angeles, Princeton University, University of Southern California, and entities like Intel, Cisco Systems, and National Science Foundation.

Overview

NDN posits that consumers request data by names instead of connecting to endpoints identified by addresses, shifting focus away from host-to-host models exemplified by ARPANET, ARP, and early DARPA designs. This approach aligns with caching strategies developed in works from Berkeley Lab and concepts illustrated in Content Delivery Network architectures pioneered at Akamai Technologies. NDN’s principles intersect with research programs funded by the European Research Council, projects at Massachusetts Institute of Technology, and testbeds managed by National Institute of Standards and Technology.

Architecture and Components

The architecture features key data-plane structures analogous to tables and caches widely used in routing systems such as Border Gateway Protocol deployments and experimental overlays at PlanetLab. Core components include a Pending Interest Table (PIT), Forwarding Information Base (FIB), and Content Store (CS), which echo design patterns from Cisco Systems routers and prototype platforms at PARC. Implementations often run on operating systems developed at Linux Foundation projects and have been integrated with middleware from Apache Software Foundation projects for evaluation. The architecture supports in-network caching like systems inspired by Akamai Technologies and forwarding strategies comparable to those studied in Google network research.

Naming and Naming Conventions

Names in this architecture are hierarchical and routable, resembling URI-like structures used across World Wide Web Consortium standards and practices from IETF working groups. Naming schemes draw influence from naming work at DNS and naming research linked to Xerox PARC publications and the IAB discussions on future routing. Researchers from Stanford University and Carnegie Mellon University have proposed human-readable and application-oriented name conventions analogous to schemes in HTTP and File Transfer Protocol. Naming also interacts with identity frameworks explored at OpenID Foundation and authentication mechanisms discussed in Internet Engineering Task Force drafts.

Routing and Forwarding Mechanisms

Routing relies on name-based lookup and stateful forwarding; algorithms build upon concepts from link-state and distance-vector protocols like Open Shortest Path First and Routing Information Protocol. Experimental routing protocols have been evaluated on testbeds such as PlanetLab and in simulation environments used by teams at MIT and University of Cambridge. Forwarding strategies leverage adaptive algorithms similar to multipath approaches researched at Bell Labs and multiprotocol label switching techniques from ITU-T studies. Interoperability experiments have involved vendors such as Juniper Networks and Ericsson to assess deployment feasibility.

Security and Trust Model

Security in this model attaches cryptographic provenance to named data objects, an approach influenced by public-key systems standardized by RSA Laboratories, National Institute of Standards and Technology, and standards like X.509. The model reuses techniques from hashing and signature schemes first formalized in works by Diffie–Hellman and Whitfield Diffie-related research, and it interfaces with trust management concepts developed at SRI International and Carnegie Mellon University. Trust schema proposals echo authorization patterns from OAuth and identity assurances studied by Microsoft Research and Google security teams. Threat modeling references adversary frameworks similar to those used in Zero Trust discussions and secure designs evaluated under Common Criteria methodologies.

Performance and Scalability

Performance evaluations compare NDN caching and stateful forwarding against traditional CDN and IP architectures used by Netflix and large-scale deployments at Facebook. Scalability studies leverage methods from high-performance research at Lawrence Berkeley National Laboratory and traffic engineering experiments inspired by AT&T backbone analyses. Simulation and emulation work often use tools developed at Georgia Institute of Technology and modeling approaches advanced in publications from IEEE conferences. Metrics include cache hit ratio, routing table size, forwarding latency, and resilience under churn as studied in research from Bell Labs Research and Microsoft Research.

Research, Implementations, and Applications

The NDN research community comprises academic labs across University of California, Berkeley, UCLA, USC, Princeton University, University of Illinois Urbana-Champaign and industry partners including Cisco Systems, Intel, and Huawei. Open-source implementations and prototypes have been developed in projects hosted by contributors from Open Networking Foundation and evaluated on testbeds like GENI and Fed4FIRE. Applications explored include content distribution for streaming comparable to YouTube workflows, Internet of Things integrations studied at ARM research, and vehicular networking trials connected to programs at Toyota Research Institute. Standardization and outreach have been discussed in venues including ACM SIGCOMM, IEEE INFOCOM, and workshops co-located with IETF meetings.

Category:Computer networking