World Library  
Flag as Inappropriate
Email this Article


Template:Infobox telecommunications network The National Science Foundation Network (NSFNET) was a program of coordinated, evolving projects sponsored by the National Science Foundation (NSF) beginning in 1985 to promote advanced research and education networking in the United States.[1] NSFNET was also the name given to several nationwide backbone networks that were constructed to support NSF's networking initiatives from 1985-1995. Initially created to link researchers to the nation's NSF-funded supercomputing centers, through further public funding and private industry partnerships it developed into a major part of the Internet backbone.


Following the deployment of the Computer Science Network (CSNET), a network that provided Internet services to academic computer science departments, in 1981, the U.S. National Science Foundation (NSF) aimed to create an academic research network facilitating access by researchers to the supercomputing centers funded by NSF in the United States.[2]

In 1985, NSF began funding the creation of five new supercomputing centers:

Also in 1985, under the leadership of Dennis Jennings, the NSF established the National Science Foundation Network (NSFNET). NSFNET was to be a general-purpose research network, a hub to connect the five supercomputing centers along with the NSF-funded National Center for Atmospheric Research (NCAR) to each other and to the regional research and education networks that would in turn connect campus networks. Using this three tier network architecture NSFNET would provide access between the supercomputer centers and other sites over the backbone network at no cost to the centers or to the regional networks using the open TCP/IP protocols initially deployed successfully on the ARPANET.

The 56-kbit/s backbone

The NSFNET initiated operations in 1986 using TCP/IP. Its six backbone sites were interconnected with leased 56-kbit/s links, built by a group including the University of Illinois National Center for Supercomputing Applications (NCSA), Cornell University Theory Center, University of Delaware, and Merit Network. PDP-11/73 minicomputers with routing and management software, called Fuzzballs, served as the network routers since they already implemented the TCP/IP standard.

This original 56-kbit/s backbone was overseen by the supercomputer centers themselves with the lead taken by Ed Krol at the University of Illinois at Urbana-Champaign. PDP-11/73 Fuzzball routers were configured and run by Hans-Werner Braun at the Merit Network[3] and statistics were collected by Cornell University.

Support for NSFNET end-users was provided by the Network Service Center (NNSC), located at BBN Technologies and included publishing the softbound "Internet Manager's Phonebook" which listed the contact information for every issued domain name and IP address in 1990.[4] Incidentally, Ed Krol also authored the Hitchhiker's Guide to the Internet to help users of the NSFNET understand its capabilities.[5] The Hitchhiker's Guide became one of the first help manuals for the Internet.

As regional networks grew the 56 K-bit/sec NSFNET backbone experienced rapid increases in network traffic and became seriously congested. In June 1987 NSF issued a new solicitation to upgrade and expand NSFNET.[6]

The 1.5 Mbit/s (T1) backbone

As a result of a November 1987 NSF award to the Merit Network, a networking consortium by public universities in Michigan, the original 56-kbit/s network was expanded to include 13 nodes interconnected at 1.5 Mbit/s (T1) by July 1988. The backbone nodes used routers based on a collection of nine IBM RT systems running AOS, IBM's version of Berkeley UNIX.

Under its cooperative agreement with NSF the Merit Network was the lead organization in a partnership that included IBM, MCI, and the State of Michigan. Merit provided overall project coordination, network design and engineering, a Network Operations Center (NOC), and information services to assist the regional networks. IBM provided equipment, software development, installation, maintenance and operations support. MCI provided the T1 data circuits at reduced rates. The state of Michigan provided funding for facilities and personnel. Eric M. Aupperle, Merit's President, was the NSFNET Project Director, and Hans-Werner Braun was Co-Principal Investigator.

From 1987 to 1994 Merit organized a series of "Regional-Techs" meetings, where technical staff from the regional networks met to discuss operational issues of common concern with each other and the Merit engineering staff.

During this period, but separate from its support for the NSFNET backbone, NSF funded:

  • the NSF Connections Program that helped colleges and universities obtain or upgrade connections to regional networks;
  • regional networks to obtain or upgrade equipment and data communications circuits;
  • the NNSC, and successor Network Information Services Manager (aka InterNIC) information help desks;[7]
  • the International Connections Manager (ICM), a task performed by Sprint, that encouraged connections between the NSFNET backbone and international research and education networks; and
  • various ad hoc grants to organizations such as the Federation of American Research Networks (FARNET).

The NSFNET became the principal Internet backbone starting in approximately 1988, when in addition to the five NSF supercomputer centers it included connectivity to the regional networks BARRNet, Merit/MichNet, MIDnet, NCAR, NorthWestNet, NYSERNet, JVNCNet, SESQUINET, SURAnet, and Westnet, which in turn connected about 170 additional networks to the NSFNET.[8] Three new nodes were added as part of the upgrade to T3: NEARNET in Cambridge, Massachusetts; Argone National Laboratory outside of Chicago; and SURAnet in Atlanta, Georgia.[9] NSFNET connected to other federal government networks including the NASA Science Internet, the Energy Science Network (ESNET), and others. Connections were also established to international research and education networks, first to France and Canada, then to NordUnet serving Denmark, Finland, Iceland, Norway, and Sweden, to Mexico, and many others.

Two Federal Internet Exchanges (FIXes) were established in June 1989[10] under the auspices of the Federal Engineering Planning Group (FEPG). FIX East, at the University of Maryland in College Park and FIX West, at the NASA Ames Research Center in Mountain View, California. The existence of NSFNET and the FIXes allowed the ARPANET to be phased out in mid-1990.[11]

Starting in August 1990 the NSFNET backbone supported the OSI Connectionless Network Protocol (CLNP) in addition to TCP/IP.[12] However, CLNP usage remained low when compared to TCP/IP.

Traffic on the network continued its rapid growth, doubling every seven months. Projections indicated that the T1 backbone would become overloaded sometime in 1990.

A critical routing technology, Border Gateway Protocol (BGP), originated during this period of Internet history. BGP allowed routers on the NSFNET backbone to differentiate routes originally learned via multiple paths. Prior to BGP, interconnection between IP network was inherently hierarchical, and careful planning was needed to avoid routing loops.[13] BGP turned the Internet into a meshed topology, moving away from the centric architecture which the ARPANET emphasized.

The 45-Mbit/s (T3) backbone

During 1991 the backbone was upgraded to 45 Mbit/s (T3) transmission speed and expanded to interconnect 16 nodes. The routers on the upgraded backbone were based on an IBM RS/6000 workstation running UNIX. Core nodes were located at MCI facilities with end nodes at the connected regional networks and supercomputing centers. Completed in November 1991, the transition from T1 to T3 did not go as smoothly as the transition from 56K to T1, took longer than planned, and as a result there was at times serious congestion on the overloaded T1 backbone. Following the transition to T3, portions of the T1 backbone were left in place to act as a backup for the new T3 backbone.

In anticipation of the T3 upgrade and the approaching end of the 5-year NSFNET cooperative agreement, in September 1990 Merit, IBM, and MCI formed Advanced Network and Services (ANS), a new non-profit corporation with a more broadly based Board of Directors than the Michigan based Merit Network. Under its cooperative agreement with NSF, Merit remained ultimately responsible for the operation of NSFNET, but subcontracted much of the engineering and operations work to ANS. Both IBM and MCI made substantial new financial and other commitments to help support the new venture. Allan Weis left IBM to become ANS's first President and Managing Director. Douglas Van Houweling, former Chair of the Merit Network Board and Vice Provost for Information Technology at the University of Michigan, was Chairman of the ANS Board of Directors.

The new T3 backbone was named ANSNet and provided the physical infrastructure used by Merit to deliver the NSFNET Backbone Service.

Regional networks

In addition to the five NSF supercomputer centers, NSFNET provided connectivity to eleven regional networks and through these networks to many smaller regional and campus networks. The NSFNET regional networks were:[9][14]

Commercial traffic

The NSF's appropriations act authorized NSF to "foster and support the development and use of computer and other scientific and engineering methods and technologies, primarily for research and education in the sciences and engineering." This allowed NSF to support NSFNET and related networking initiatives, but only to the extent that that support was "primarily for research and education in the sciences and engineering."[18] And this in turn was taken to mean that use of NSFNET for commercial purposes was not allowed.

The NSFNET Backbone Services Acceptable Use Policy[19]
June 1992

         General Principle
  1. NSFNET Backbone services are provided to support open research and education in and among US research and instructional institutions, plus research arms of for-profit firms when engaged in open scholarly communication and research. Use for other purposes is not acceptable.
  2. Specifically Acceptable Uses

  3. Communication with foreign researchers and educators in connection with research or instruction, as long as any network that the foreign user employs for such communication provides reciprocal access to US researchers and educators.
  4. Communication and exchange for professional development, to maintain currency, or to debate issues in a field or subfield of knowledge.
  5. Use for disciplinary-society, university-association, government-advisory, or standards activities related to the user's research and instructional activities.
  6. Use in applying for or administering grants or contracts for research or instruction, but not for other fundraising or public relations activities.
  7. Any other administrative communications or activities in direct support of research and instruction.
  8. Announcements of new products or services for use in research or instruction, but not advertising of any kind.
  9. Any traffic originating from a network of another member agency of the Federal Networking Council if the traffic meets the acceptable use policy of that agency.
  10. Communication incidental to otherwise acceptable use, except for illegal or specifically unacceptable use.
  11. Unacceptable Uses

  12. Use for for-profit activities, unless covered by the General Principle or as a specifically acceptable use.
  13. Extensive use for private or personal business.

This statement applies to use of the NSFNET Backbone only. NSF expects that connecting networks will formulate their own use policies. The NSF Division of Networking and Communications Research and Infrastructure will resolve any questions about this Policy or its interpretation.

Acceptable Use Policy (AUP)

To ensure that NSF support was used appropriately, NSF developed an NSFNET Acceptable Use Policy (AUP) that outlined in broad terms the uses of NSFNET that were and were not allowed.[19] The AUP was revised several times to make it clearer and to allow the broadest possible use of NSFNET, consistent with Congress' wishes as expressed in the appropriations act.

A notable feature of the AUP is that it talks about acceptable uses of the network that are not directly related to who or what type of organization is making that use. Use from for-profit organizations is acceptable when it is in support of open research and education. And some uses such as fundraising, advertising, public relations activities, extensive personal or private use, for-profit consulting, and all illegal activities are never acceptable, even when that use is by a non-profit college, university, K-12 school, or library. And while these AUP provisions seem quite reasonable, in specific cases they often proved difficult to interpret and enforce. NSF did not monitor the content of traffic that was sent over NSFNET or actively police the use of the network. And it did not require Merit or the regional networks to do so. NSF, Merit, and the regional networks did investigate possible cases of inappropriate use, when such use was brought to their attention.[20]

An example may help to illustrate the problem. Is it acceptable for a parent to exchange e-mail with a child enrolled at a college or university, if that exchange uses the NSFNET backbone? It would be acceptable, if the subject of the e-mail was the student's instruction or a research project. Even if the subject was not instruction or research, the e-mail still might be acceptable as private or personal business as long as the use was not extensive.[21]

The prohibition on commercial use of the NSFNET backbone[22] meant that some organizations could not connect to the Internet via regional networks that were connected to the NSFNET backbone, while to be fully connected other organizations (or regional networks on their behalf), including some non-profit research and educational institutions, would need to obtain two connections, one to an NSFNET attached regional network and one to a non-NSFNET attached network provider. In either case the situation was confusing and inefficient. It prevented economies of scale, increased costs, or both. And this slowed the growth of the Internet and its adoption by new classes of users, something no one was happy about.

Commercial ISPs, ANS CO+RE, and the CIX

During the period when NSFNET was being established, Internet service providers that allowed commercial traffic began to emerge, such as Alternet, PSINet, CERFNet, and others. The commercial networks in many cases were interconnected to the NSFNET and routed traffic over the NSFNET nominally accordingly to the NSFNET acceptable use policy[23] Additionally, these early commercial networks often directly interconnected with each other as well as, on a limited basis, with some of the regional Internet networks.

In 1991, the Commercial Internet eXchange (CIX, pronounced "kicks") was created by PSINet, UUNET and CERFnet to provide a location at which multiple networks could exchange traffic free from traffic-based settlements and restrictions imposed by an acceptable use policy.[24]

In 1991 a new ISP, ANS CO+RE (commercial plus research), raised concerns and unique questions regarding commercial and non-commercial interoperability policies. ANS CO+RE was the for-profit subsidiary of the non-profit Advanced Network and Services (ANS) that had been created earlier by the NSFNET partners, Merit, IBM, and MCI.[25] ANS CO+RE was created specifically to allow commercial traffic on ANSNet without jeopardizing its parent's non-profit status or violating any tax laws. The NSFNET Backbone Service and ANS CO+RE both used and shared the common ANSNet infrastructure. NSF agreed to allow ANS CO+RE to carry commercial traffic subject to several conditions:

  • that the NSFNET Backbone Service was not diminished;
  • that ANS CO+RE recovered at least the average cost of the commercial traffic traversing the network; and
  • that any excess revenues recovered above the cost of carrying the commercial traffic would be placed into an infrastructure pool to be distributed by an allocation committee broadly representative of the networking community to enhance and extend national and regional networking infrastructure and support.

For a time ANS CO+RE refused to connect to the CIX and the CIX refused to purchase a connection to ANS CO+RE. In May 1992 Mitch Kapor and Al Weis forged an agreement where ANS would connect to the CIX as a "trial" with the ability to disconnect at a moment's notice and without the need to join the CIX as a member.[26] This compromise resolved things for a time, but later the CIX started to block access from regional networks that had not paid the $10,000 fee to become members of the CIX.[27]

An unfortunate state of affairs

The creation of ANS CO+RE and its initial refusal to connect to the CIX was one of the factors that lead to the controversy described later in this article. Other issues had to do with:

  • differences in the cultures of the non-profit research and education community and the for-profit community with ANS trying to be a member of both camps and not being fully accepted by either;
  • differences of opinion about the best approach to take to open the Internet to commercial use and to maintain and encourage a fully interconnected Internet; and
  • differences of opinion about the correct type and level of involvement in Internet networking initiatives by the public and the private sectors.

For a time this unfortunate state of affairs kept the networking community as a whole from fully implementing the true vision for the Internet—a world-wide network of fully interconnected TCP/IP networks allowing any connected site to communicate with any other connected site. These problems would not be fully resolved until a new network architecture was developed and the NSFNET Backbone Service was turned off in 1995.

Privatization and a new network architecture

The NSFNET Backbone Service was primarily used by academic and educational entities, and was a transitional network bridging the era of the ARPANET and CSNET into the modern Internet of today.

On April 30, 1995, the NSFNET Backbone Service had been successfully transitioned to a new architecture[28] and the NSFNET backbone was decommissioned.[29] At this point there were still NSFNET programs, but there was no longer an NSFNET network or network service.

After the transition, network traffic was carried on any of several commercial backbone networks, internetMCI, PSINet, SprintLink, ANSNet, and others. Traffic between networks was exchanged at four Network Access Points or NAPs. The NAPs were located in New York (actually New Jersey), Washington, D.C., Chicago, and San Jose and run by Sprint, MFS Datanet, Ameritech, and Pacific Bell.[30] The NAPs were the forerunners of modern Internet exchange points.

The former NSFNET regional networks could connect to any of the new backbone networks or directly to the NAPs, but in either case they would need to pay for their own connections. NSF provided some funding for the NAPs and interim funding to help the regional networks make the transition, but did not fund the new backbone networks directly.

To help ensure the stability of the Internet during and immediately after the transition from NSFNET, NSF conducted a solicitation to select a Routing Arbiter (RA) and ultimately made a joint award to the Merit Network and USC's Information Science Institute to act as the RA.

To continue its promotion of advanced networking technology the NSF conducted a solicitation to create a very high-speed Backbone Network Service (vBNS) that, like NSFNET before it, would focus on providing service to the research and education community. MCI won this award and created a 155 M-bit/sec (OC3c) and later a 622 M-bit/sec (OC12c) and 2.5 G-bit/sec (OC48c) ATM network to carry TCP/IP traffic primarily between the supercomputing centers and their users. NSF support[31] was available to organizations that could demonstrate a need for very high speed networking capabilities and wished to connect to the vBNS or to the Abilene Network, the high speed network operated by the University Corporation for Advanced Internet Development (UCAID, aka Internet2).[32]

At the February 1994 regional techs meeting in San Diego, the group revised its charter[33] to include a broader base of network service providers, and subsequently adopted North American Network Operators' Group (NANOG) as its new name. Elise Gerich and Mark Knopper were the founders of NANOG and its first coordinators, followed by Bill Norton, Craig Labovitz, and Susan Harris.[34]


For much of the period from 1987 to 1995, following the opening up of the Internet through NSFNET and in particular after the creation of the for-profit ANS CO+RE in May 1991, some Internet stakeholders[35] were concerned over the effects of privatization and the manner in which ANS, IBM, and MCI received a perceived competitive advantage in leveraging federal research money to gain ground in fields in which other companies allegedly were more competitive. The Cook Report on the Internet,[36] which still exists, evolved as one of its largest critics. Other writers, such as Chetly Zarko, a University of Michigan alumnus and freelance investigative writer, offered their own critiques.[37]

On March 12, 1992 the Subcommittee on Science of the Committee on Science, Space, and Technology, U.S. House of Representatives, held a hearing to review the management of NSFNET.[20] Witnesses at the hearing were asked to focus on the agreement(s) that NSF put in place for the operation of the NSFNET backbone, the foundation's plan for recompetition of those agreements, and to help the subcommittee explore whether the NSF's policies provided a level playing field for network service providers, ensured that the network was responsive to user needs, and provided for effective network management. The subcommittee heard from seven witnesses, asked them a number of questions, and received written statements from all seven as well as from three others. At the end of the hearing, speaking to the two witnesses from NSF, Dr. Nico Habermann, Assistant NSF Director for the Computer and Information Science and Engineering Directorate (CISE), and Dr. Stephen Wolff, Director of NSF's Division of Networking & Communications Research & Infrastructure (DNCRI), Representative Boucher, Chairman of the subcommittee, said:

"… I think you should be very proud of what you have accomplished. Even those who have some constructive criticism of the way that the network is presently managed acknowledge at the outset that you have done a terrific job in accomplishing the goal of this NSFNET, and its user-ship is enormously up, its cost to the users has come down, and you certainly have our congratulations for that excellent success."

Subsequently the subcommittee drafted legislation, becoming law on October 23, 1992, which authorized the National Science Foundation

… to foster and support access by the research and education communities to computer networks which may be used substantially for purposes in addition to research and education in the sciences and engineering, if the additional uses will tend to increase the overall capabilities of the networks to support such research and education activities (that is to say, commercial traffic).[38]

This legislation allowed, but did not require, NSF to repeal or modify its existing NSFNET Acceptable Use Policy (AUP)[19] which restricted network use to activities in support of research and education.[22]

The hearing also led to a request from Rep. Boucher asking the NSF Inspector General to conduct a review of NSF's administration of NSFNET. The NSF Office of the Inspector General released its report on March 23, 1993.[25] The report concluded by:

  • stating that "[i]n general we were favorably impressed with the NSFNET program and staff";
  • finding no serious problems with the administration, management, and use of the NSFNET Backbone Service;
  • complimenting the NSFNET partners, saying that "the exchange of views among NSF, the NSFNET provider (Merit/ANS), and the users of NSFNET [via a bulletin board system], is truly remarkable in a program of the federal government"; and
  • making 17 "recommendations to correct certain deficiencies and strengthen the upcoming re-solicitation."


External links

  • The Internet - the Launch of NSFNET, National Science Foundation
  • NSFNET: A Partnership for High-Speed Networking, Final Report 1987-1995, Karen D. Frazer, Merit Network, Inc., 1995
  • NSF and the Birth of the Internet, National Science Foundation, December 2007
  • San Diego State University
  • "Fool Us Once Shame on You—Fool Us Twice Shame on Us: What We Can Learn from the Privatizations of the Internet Backbone Network and the Domain Name System", Jay P. Kesan and Rajiv C. Shah, Washington University Law Review, Volume 79, Issue 1 (2001)
  • centennial year
  • Merit Network: A history
  • Finnish IT center for science
    • Hathi Trust Digital Library
  • Reflection on NSFNet

Template:American research and education networks

This article was sourced from Creative Commons Attribution-ShareAlike License; additional terms may apply. World Heritage Encyclopedia content is assembled from numerous content providers, Open Access Publishing, and in compliance with The Fair Access to Science and Technology Research Act (FASTR), Wikimedia Foundation, Inc., Public Library of Science, The Encyclopedia of Life, Open Book Publishers (OBP), PubMed, U.S. National Library of Medicine, National Center for Biotechnology Information, U.S. National Library of Medicine, National Institutes of Health (NIH), U.S. Department of Health & Human Services, and, which sources content from all federal, state, local, tribal, and territorial government publication portals (.gov, .mil, .edu). Funding for and content contributors is made possible from the U.S. Congress, E-Government Act of 2002.
Crowd sourced content that is contributed to World Heritage Encyclopedia is peer reviewed and edited by our editorial staff to ensure quality scholarly research articles.
By using this site, you agree to the Terms of Use and Privacy Policy. World Heritage Encyclopedia™ is a registered trademark of the World Public Library Association, a non-profit organization.

Copyright © World Library Foundation. All rights reserved. eBooks from World Library are sponsored by the World Library Foundation,
a 501c(4) Member's Support Non-Profit Organization, and is NOT affiliated with any governmental agency or department.