Return-Path: <[email protected]>
Date: Fri, 13 Mar 1992 10:31:31 -0800 (PST)
From: Eric Hood {NWNet} <[email protected]>
Sender: Eric Hood {NWNet} <[email protected]>
Reply-To: Eric Hood {NWNet} <[email protected]>
Subject: Eric Hood's Testimony to House Science Subcommittee
To: [email protected]
Cc: Eric Hood {NWNet} <[email protected]>
Testimony of Dr. Eric S. Hood, President, Federation of American Research
Networks (FARNET), and Executive Director, NorthWestNet, as submitted to
the U.S. House of Representatives Subcommittee on Science of the Committee
on Science, Space, and Technology

Hearing on the Operation, Management, and Future of the National Science
Foundation Network (NSFNET) and the National Research and Education
Network (NREN)

March 12, 1992

Chairman Boucher and members of the Committee, I am truly pleased to
present testimony regarding the National Science Foundation Network
(NSFNET) and the National Research and Education Network (NREN) on
behalf of the Federation of American Research Networks (FARNET).

FARNET is a non-profit association chartered to support the evolution
and wide-spread adoption of data networking to enhance research,
education, and economic development.  Our 32 members include
operators of state, regional, and national computer networks,
telecommunications vendors, and other organizations of like mission.
A majority of the private network providers [e.g., Advanced Network
Services (ANS), California Education and Research Foundation Network
(CERFnet), and Performance Systems International (PSI)] are members,
as are most of the other regional and state public networks connected
to the National Science Foundation Network (NSFNET).  FARNET also has
liaison relationships with other national and international
organizations with interests in networking, including the
Coordinating Committee on Intercontinental Research Networks (CCIRN)
and the Internet Engineering Task Force (IETF).

As a membership association, FARNET limits its public commentary on
policy to questions which the membership has considered as a body.
With regard to the issues before the Subcommittee, and in response to
your questions regarding the NSFNET and the National Research and
Education Network (NREN) as posed in your letter of February 28,
1992, the FARNET Board of Directors would like to communicate the
following observations based on workshops, conferences, and focused
electronic discussions conducted over the past six months.

"Assessment of NSF's efforts to provide networking support to the
science and engineering research and education community, including
your views on the current arrangement for operation of NSFNET."

The Division of Networking and Communications Research and
Infrastructure (DNCRI) at NSF has done an exemplary job of enabling
the provision of network access to the nation's research and
education communities.  Over the past five years, NSF has created a
nd maintained an operational infrastructure which has successfully
sustained exponential growth, as measured by the number of users, by
the number of connected institutions, and by the volume of network
traffic.  In addition, NSF has demonstrated technical leadership in
establishing standards for data networking and for supporting
research into very high speed technologies.  Furthermore, NSF's
investments, which have been modest by Federal standards, have
leveraged significant funding commitments on university campuses,
>from corporations, and from state governments.

Ten years ago in 1982, the Defense Advanced Research Project Agency's
ARPANET deployed then state of the art technology to connect military
facilities to approximately 10 or 15 research universities.  The
computing devices attached to the network numbered in the hundreds
and ARPANET trunk speeds did not exceed 56 thousand bits per second.
The custom-built hardware needed to accomplish the switching cost on
the order of $100,000 to $200,000 per node.  Computer scientists and
engineers were the only regular users of the network.

In contrast, trunk bandwidths in the NSFNET now routinely exceed 25
times the capacity of the original ARPANET (or 1.5 million bits per
second), and many trunks are capable of supporting 700 times this
amount (or 45 million bits per second).  The cost of the switching
hardware is at least one, and can be two, orders of magnitude
smaller.  The computing devices that we use daily, from very fast
personal computers to graphics workstations to parallel
supercomputers, were barely envisioned by designers 10 ye ars ago.

Today's NSFNET allows researchers and scholars to communicate
electronically, to exchange text, data and graphical images, and to
access geographically dispersed information resources.  The NSFNET
now connects over 630 colleges and universities, or approximately 35
percent of our nation's four year institutions of higher education.
Over 90 percent of the nation's Federally sponsored research is
conducted at institutions of higher education connected to the
NSFNET.  Approximately 70 percent of our nation' s student population
attending four year colleges and universities have institutional
access to the NSFNET.  This communications and information
infrastructure enables computer users at sites across the nation to
share information and to work collaboratively on common tasks and
projects.  More than 1,000 institutions, including colleges,
universities, and not-for-profit, government, and corporate research
facilities representing every state, are currently connected to the
NSFNET.  Today's NSFNET is also an important part of a larger
communications network, the global Internet, which connects an
estimated 750,000 computers and 5 million users worldwide.

Such broad interconnectivity between our nation's colleges and
universities has opened new avenues for communication among the
nation's scholars.  The NSFNET now empowers researchers and educators
at more than just the Carnegie research universities.  Over the
network, faculty and students at comprehensive and liberal arts
colleges now routinely collaborate electronically with their
counterparts at our nation's elite research universities.

Over the past five years, NSF has provided national administrative
and technical leadership in the field of research and education
networking.  Throughout this period, NSF has provided clear and
consistent technical direction promoting the deployment of
interoperable and open data communications technologies.  The NSF has
enabled the development of a national networking infrastructure
capable of supporting communications across diverse computing
platforms manufactured by multiple vendors.  For those that have
followed NSF's lead, inter-institutional connectivity across campus,
corporate, state, regional, and national boundaries is now possible.
Thus NSF has successfully leveraged technically sound investments on
many university and corporate campuses in anticipation of national
and global interconnectivity.  One regional network, WESTnet,
estimates that for each Federal dollar invested in the NSFNET
program, the Federal Government leverages thirty-one dollars of state
and local funds.

"Regarding NSF's plan for recompetition of the award for operation of
the NSFNET backbone: Did the plan take into account the views of the
network user and network provider communities?"

In constructing the plan for recompetition of the award for operation
of the NSFNET backbone, NSF both solicited and implemented
recommendations from network users and network providers represented
by FARNET.  With support from the National Science Foundation,
FARNET conducted a workshop and electronic discussion in the late
summer and early fall of 1991 to address the complex issue of
inter-regional connectivity.  Specifically, FARNET considered future
options for the provision of connectivity among midlevel networks
after the current agreement for NSFNET backbone services expires in
November, 1992.  Participants included representatives from FARNET
member networks, other NREN stake holders (including
telecommunications carriers and leaders in university information
technology), Federal agency representatives, and legal and economic
experts.  Our report was well received by NSF.  In fact, several key
concepts from the FARNET report regarding network stability and
multiple awards to inter-regional connectivity service providers
were included in NSF's report to the National Science Board, "Project
Development Plan: Continuation and Enhancement of NSFNET Backbone
Services." (A copy of FARNET's Recommendations to the National
Science Foundation Regarding Inter-regional Connectivity is
attached.)

"More specifically, will the plan lead to a level playing field for
the commercial network providers?"

In their report to the National Science Board, "Project Development
Plan: Continuation and Enhancement of NSFNET Backbone Services," NSF
identified the two seminal issues associated with the recompetition
process: preservation of network stability and pro motion of
competition.  Regarding fair competition, NSF's report further
identifies two essential points: that the incumbent provider is not
favored, and that equal opportunity is provided to other firms
desiring to participate in the provision of transcontinental TCP/IP
networking services.

To facilitate the accomplishment of these objectives, the NREN
Engineering Group (NEG) advising the National Science Foundation has
proposed clear separation of the administration of network routing
>from the provision of transcontinental circuits and digital
switching fabric.  Operationally, this separation will be achieved
through two independent solicitations.  The NEG has recommended that
the solicitation providing for the formation of an Internet Routing
Authority (IRA) be awarded to a single entity.  FARNET agrees that
this single award is necessary to ensure the continued operational
viability of the network.  The solicitation for connectivity services
will include the possibility for multiple awards.  FARNET agrees and
has strongly recommended that the provision of connectivity services
be awarded to at least two competing providers.  Thus all awardees
will have appropriate incentives to cooperate with each other in the
development and operation of interconnection facilities.

If implemented as outlined in the report to the National Science
Board, the NSF plan for recompetition of the award for operation of
the NSFNET backbone will indeed level the playing field for the
commercial network providers while preserving an acceptable level of
stability.

"Are there better alternatives to the proposed plan?"

The plan as presented in NSF's report to the National Science Board,
"Project Development Plan: Continuation and Enhancement of NSFNET
Backbone Services," is consistent with the consensus position of a
majority of FARNET's constituency.  FARNET endorses the plan as
presented, but must again emphasize the importance of multiple awards
for the provision of inter-regional connectivity services.

"What are your views on the key issues Congress needs to consider to
help ensure a successful evolution of the current Internet to the
NREN?  What is your vision for the NREN and how would you define the
roles of the public and private sectors in realizing that vision?
What specific steps should be taken by the Congress and the Federal
agencies to help ensure the goals for the NREN are achieved?"

Something significant is happening in our country.  Over the past
five years under NSF's leadership, America has made great progress
toward creating a national data communications and information
infrastructure to enable research, education, technology transfer,
and economic development.  This national resource is already
delivering material benefit to research and education, both in the
public and private sectors.  Fully realized, the NREN holds the
promise of significantly enhancing our national competitiveness in
the global marketplace.

Over the past five years, the Federal Government has played a pivotal
role in the creation, growth, and evolution of the NSFNET.  Continued
Federal investments in the NREN must be focused to ensure equal and
ubiquitous access to our nation's information resources, to improve
network reliability, performance, and usability, and to enable
research in new and pre-competitive technologies.  With judicious
investments and the correct partnership among government, academia,
and industry, we can realize the NREN vision.

FARNET endorses the right of equal access to our nation's information
resources.

The NREN has the potential to reduce traditional impediments to data
communication and access to information systems: geographic
isolation, smallness of size, and sparseness of local resources.
Through the NSFNET this vision has become reality for our re search
and higher education communities interested in information
exploration.  From small town doctors collaborating with inner city
clinicians on a difficult diagnosis, to patrons of a community
library wishing to broaden their understanding of the world around
them, the NREN can begin to deliver on this promise by funding
programs aimed at enabling these previously disenfranchised
communities.

Through the NREN, this vision can become reality for all Americans.
The benefits provided by the NREN are not restricted to those who are
resource poor, but are shared equally by stewards of unique
resources.  For example, the Smithsonian Astrophysical Center's
collection of astronomical images in Boston and the globally
distributed antennas of the National Radio Astronomy Observatory are
equally accessible to scholars through the NSFNET.  The evolution of
the NREN must be guided by the principle of equality of access to
scholarly information for all Americans.

There is considerable historical precedent for Federal leadership in
supporting the creation, operation, and maintenance of the nation's
commerce, transportation, and utility systems.  In that spirit,
FARNET recommends that Congress continue to support the extension of
basic connectivity services to underserved and geographically remote
communities through the programs of the Federal agencies
participating in the NREN.

FARNET endorses the principles of ubiquitous access to and universal
interconnectivity of our nation's information resources.

The strength of the Internet, the system of networks that includes
the NSFNET, is its broad interconnectivity.  Internet technology is
now widely employed to electronically link computers -- from
microcomputers to supercomputers.  These computers can be located
within an organization at a single site or within multiple
organizations at opposite ends of the country.

We are well on the way to creating a national, and in the near
future, even global, communication and information network.  Any
decision that limits interconnectivity among segments of the Internet
impedes the development and reduces the usefulness of this national
resource.  Educators, researchers, and students must have access to
the full array of computing and information resources on the
Internet.  Balkanization of the Internet cannot be tolerated.
Congress must discourage funding policies which allow procurement of
network services from disjoint islands of service providers.
Endorsing procurement criteria that require the service providers to
guarantee interconnectivity to other service providers will ensure
that we progress toward a national data communications fabric that
reaches every campus, every library, every school, every home.

FARNET endorses continued efforts to improve the reliability,
performance, and usability of the networking infrastructure.

Continuing enhancements of network capacity and performance, and
improvements in network operations and engineering practices are
essential to the development of a reliable and robust NREN.
Providing an adequate level of stability and predictability, both in
the operation of the current Internet and in the transition to new
technologies or management paradigms, is critical to the continuing
growth and use of the NREN.  Until these advances in network capacity
and performance occur as a natural consequencies of private sector
activity, judicious Federal investments in the nation's
communications infrastructure will be required.  The continued
investment of Federal funds will ensure that major segments of our
population (i.e., remote areas, underserved communities,
disadvantaged constituencies) are not disenfranchised from this
national resource.

Greatly improved support for user and information services
network-wide will enable the rapid extension, acceptance, and use of
the NREN.  This support should include plans for the provision of
access to both public and private information resources, with early
resolution of copyright and other intellectual property issues.
Federal investment in research and development of directory services,
network navigation tools, user documentation, and training services
is a necessary prerequisite to enhancing network usability.

FARNET endorses the elimination of restrictions on the delivery of
commercial services across the NSFNET and the emerging NREN.

We have reached a crossroads in the evolution of true network
interconnectivity for our country.  The network infrastructure that
we are building has the inherent capability to promote research,
education, technology transfer, and economic development.  Yet, the
current acceptable use restrictions on the Federally sponsored
national backbone impede our progress.

Close collaboration, even partnership, among government, academia,
and industry is essential in promoting the advancement of these
critical technologies.  Such collaborative efforts are enabled when
all partners can, without restriction, access common infrastructure.
Yet, current policies prohibit the use of the Federally funded
network for commercial purposes.  It is as if we were required to
have two telephones on every desk, one for making purely
"educational" calls and the second for making "commercial" calls.
The economic inefficiencies and practical disadvantages of such a
system are apparent.  If the NREN is to realize its full potential,
part of its evolution must include controlled experimentation
permitting interconnection of the public and private sectors to
achieve a "critical mass" of network users and suppliers.

To remedy the current limitations on interconnectivity between the
public and private sectors, FARNET requests that Congress act to
remove the current acceptable use restrictions on the Federally
sponsored national network infrastructure.  At a minimum, commercial
traffic should be permitted on this infrastructure on an experimental
basis and under suitably controlled conditions.  This experiment
should be designed to produce results which can be analyzed from
multiple perspectives (i.e., technical, administrative, economic,
and legal) and should be widely disseminated as envisioned in the
High Performance Computing and Communications (HPCC) legislation.

FARNET endorses the principle of inter-agency cooperation and
collaboration in the construction of the NREN.

As noted earlier, one of the great strengths of today's Internet is
the provision of broad interconnectivity.  Although this
interconnectivity includes the current NSFNET and the national
mission agency networks operated by the Department of Energy (DOE),
the National Aeronautics and Space Administration (NASA), and the
Defense Advanced Research Projects Agency (DARPA), today's Internet
can not be characterized as a single, homogeneous entity.  With such
a broad and varied constituency, the Interim Interagency NREN is
unlikely to evolve as a single, homogeneous initiative, at least in
the short term.  Yet insofar as existing technologies and current
deployment strategies are sufficient to adequately support mission
agency activities, the Federal agencies must be encouraged to
collaborate and cooperate to reduce costs and to promote the public
good.

FARNET endorses further research into key technologies.

The nation must have focused research programs pushing the envelope
of high performance production networking.  Continued Federal support
for basic research is essential to the solution of difficult problems
that remain in several critical areas (e.g., network security,
authentication, privacy, routing and addressing, high speed circuit
and switching architectures).

Forging the appropriate mix of private investment and public
sponsorship is critical to the continuing development and delivery of
this important technology.  Continued Federal investments in these
key research areas will leverage private funds and enable the transfer
of pre-competitive technologies from academic and government
laboratories to private industry.

FARNET endorses the continued close collaboration among government,
academia, and industry to realize the NREN vision.

Ten years ago, the Internet was a government-funded research project.
Today, internetworking is a multi-billion dollar industry.

Ten years ago, access to this technology was limited to a small cadre
of experts in computer science and telecommunications engineering.
Today, an estimated 5 million researchers and educators, teachers and
students, authors and librarians, physicians and clinicians, policy
makers and corporate planners use the worldwide Internet.

Ten years ago, Internet devices were the constructs of research labs.
Today, Internetworking hardware and software components are
commercially available from a variety of vendors at commodity prices.

Ten years ago, telecommunications providers were offering dedicated
digital services within and between only a few metropolitan areas.
Today, every major provider offers, or has immediate plans to offer
tariffed, high speed packetized digital services aim ed at a
burgeoning data networking market.

>From these achievements it is clear that we have made considerable
progress along the sometimes bumpy road toward commercialization and
privatization of the Internet.  The components of networking
technology have matured and become available commercially. Prices of
networking equipment, circuits, and services have dropped for all
consumers.  New vendors with new capital have been attracted into the
market, creating new high-technology jobs and opportunities.

But the internetworking industry faces the classical set of
challenges associated with success management.  While business
activity in the private sector grows, we must avoid the temptation to
reduce government investment in new or pre-competitive technologies.
While the costs of connectivity in our metropolitan areas continue to
decrease, we must avoid the temptation to reduce government support
for communities that are underserved.

If we move cautiously forward in the correct partnership of
government, academia, and industry, our nation can work together to
realize the NREN vision.

We trust that these observations will be of use to the Subcommittee
in its March deliberations.  These comments are respectfully
submitted by Dr. Eric S. Hood, President of the Federation of
American Research Networks (FARNET).

You are viewing proxied material from infinitelyremote.com. The copyright of proxied material belongs to its original authors. Any comments or complaints in relation to proxied material should be directed to the original authors of the content concerned. Please see the disclaimer for more details.