|
The Next Generation Internet Research Act of 1998,
requires the President's Information Technology (IT) Advisory Committee
(PITAC) to review the implementation of the Next Generation Internet
(NGI) initiative and report annually on:
| |
Advanced Networking Research: Progress in NGI-funded
advanced networking research, NGI Testbeds: Progress
in implementing high-performance network testbeds,
NGI Applications: Progress in developing high-performance
network applications,
Geographic Reach: Addressing geographic penalties faced
by rural institutions,
Minority- and Small-College Reach: Addressing access
by historically black and Hispanic-serving institutions, colleges
and universities with fewer than 5,000 students,
Technology Transfer: Flow of NGI ideas to industry,
Agency Coordination: Effectiveness of coordination among the
NGI agencies, and
IT Leadership: The extent to which Federal research
support will maintain U.S. IT leadership .
|
The Committee reviewed the NGI initiative with DARPA, DOE, NASA,
NIH, NIST, and NSF.
The NGI agencies have made excellent progress over the last year.
We are gratified that they have responded to advice from Congress
and from the PITAC. We recommend Congress extend funding of the
NGI program for two more years at the proposed level of $104 million.
Advanced networking research: The NGI program has initiated
a broad networking research plan that addresses the need for critical
technology. NGI research activities, especially the DARPA and NSF
programs, are producing exciting results that are likely to be cornerstones
of the new communications infrastructure. For example, DARPA's SuperNet
program will provide guaranteed ultra high-bandwidth on demand for
national, regional, metropolitan, and local networks. They are pioneering
work on dynamically controlled optical networks.
NGI Testbeds: NSF and DARPA are deploying two NGI Testbeds.
In FY 2000, NSF, DOE, and NASA have connected more than 150 sites
to a testbed providing a 100-fold increase in network performance,
and DARPA is deploying a testbed with 1000-fold increased performance
at more than 15 sites to support networking research and applications
development. Currently, the maximum desktop-to-desktop performance
has been measured at about 900 million bits per second more than
a ten-fold increase over the 80 Mbps reported to us last year. There
is still much more to do, but the agencies have an excellent plan
and they are executing it very well. The agencies have also cooperated
in setting up the Qbone network to experiment with applications
that need Quality of Service guarantees.
NGI Applications: The NGI agencies are exploring network
applications that require NGI services and performance. NIH's National
Library of Medicine, for example, has funded many promising application
pre-proposals. Many agencies (NASA, DARPA, USGS, NOAA, NEMA, and
ACE) are cooperating on the Digital Earth framework to make all
US geo-referenced data available via a point-and-click user interface
via a high-speed network. Today, there are few applications that
will use the low latency, quality of service, or high bandwidth
that NGI will offer. Only 20 percent of NGI's FY 2000 funding is
for applications, hence progress has been disappointing. We again
recommend increased emphasis on and investments in new applications
that use NGI-level capabilities. It is necessary to fund such
applications explicitly. Congress might consider doing this as part
of other agency funding bills.
Reach: The NGI initiative cannot directly address reach
to rural, inner-city, minority, or small institutions. NGI is a
research and development program to provide the technologies and
applications required as foundations for the next generations of
the Internet. It is a relatively small initiative that funds peer-reviewed
research proposals. The announced NGI awards cover 177 of the nation's
2,200 four-year, college-level educational institutions. It cannot
fund institutions where research is not emphasized and where there
is little experience developing advanced networking or applications.
The end of this report makes a specific recommendation to address
this issue.
| |
Geographic Reach: While the NGI initiative was not
planned or directed to address reach in general, access for
otherwise qualified universities with fundable research proposals
must not be disadvantaged merely because of their location.
NSF expanded the High Performance Connections program to cover
all 50 states and has made 40 grants in 18 EPSCoR states.
Eventually, NGI research on wireless, hybrid, and satellite
technologies may reduce the cost and improve the services
available to all users including those in geographically remote
areas.
Minority- and Small-College Reach: The NGI was not
funded to address Internet access for historically black,
Hispanic-serving, Native American, or small colleges and universities.
Based on its standard peer review process, NSF has awarded
High Performance Connection grants to two historically black
and five Hispanic-serving institutions. In addition, unrelated
to the NGI effort, NSF has granted 6 million dollars to Educause
to work with disadvantaged minority colleges to develop their
IT infrastructure.
|
Technology Transfer: Industry, universities, and Federal
agencies collaborate on NGI research, testbeds, teams, and workshops.
This creates substantial technology transfer opportunities. Indeed,
a quick survey found about a dozen startups capitalized at nearly
$30 billion that have sprung up from the NGI program.
Agency Coordination: NGI agencies coordinate their activities
through the Large Scale Networking Coordinating Group and its specialized
teams that focus on the NGI goals. The agencies seem to be cooperating
effectively in implementing the NGI goals.
IT Leadership: The Committee's Report to the President,
February 1999, concluded that Federal support for research in information
technology is seriously inadequate. The PITAC recommended the Federal
government increase IT research funding by $1.4B per year by FY2004.
The President's FY 2000 Budget includes an Information Technology
for the 21st Century (IT2) initiative that begins to address this
recommendation.
Recommendations
The Committee recommends that the NGI program be extended
through FY 2002 at the proposed funding levels. Planning
for follow-on activities should begin now. In particular,
more applications should be funded that demonstrate the
utility of the NGI's gigabit bandwidth to end-users, its
increased security, and its expanded quality of service.
The Committee shares Congress' concern that no Federal
program addresses the reach issue. We recommend Congress
consider additional funding for a program where the NGI
research institutions act as aggregators and mentors for
nearby smaller or disadvantaged institutions. This is
primarily infrastructure, not networking research and, hence,
not part of the current NGI program.
|
| |
Raj Reddy
Co-Chair PITAC |
Irving Wladawsky-Berger
Co-Chair PITAC |
|
