|
Items in cart [1] |
Questions? Call 888-624-8373 |
PAPERBACK + PDF PAPERBACK PDF BOOK PDF CHAPTERS Free Resources |
||||||||||||||||
|
||||||||||||||||
|
|
|||||||||||||||
| ||||||||||||||||||||||||||||||
|
||||||||||||||||||||||||||||||
| Copyright © 2009. National Academy of Sciences. All rights reserved. Terms of Use and Privacy Statement |
Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter.
Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 29
Report of the Pane! on the Network
Systems and Communications Industry
The Panel on Network Systems and Communications, one of five panels
formed by the Committee on the Impact of Academic Research on Industrial
Performance, was asked to examine the impact of academic research on the
performance of the network systems and communications industry and recom-
mend ways based on trends in the industry and the research community to in-
crease this impact. The panel of six included three members of NAB (all from
industry), one other member from industry, and two from academia. Three of the
panel members were also members of the parent committee. The panel reviewed
the literature, developed several case studies, and sent a questionnaire to experts
in academia, the computer industry, the network systems and communications
industry, and government. The questionnaire was followed by a workshop at-
tended by approximately 30 senior individuals in the network systems and com-
munications sector (see Addendum).
The network systems and communications business sector flourished
throughout the l990s, when the growth of the Internet, the technologies that
implement it, and the businesses and services that use it were unprecedented.
Telecommunications services especially wireless digital telephones and paging
services also grew rapidly. Much of this success was attributable to exponential
improvements in the performance-to-cost ratio of microelectronics over the past
three decades. Technical innovations emerging from within the industry and from
academic research have been essential. Some innovations were the culmination
of decades of research; some were short-term developments that entered the
29
OCR for page 30
30 THE IMPACT OF ACADEMIC RESEARCH ON INDUSTRIAL PERFORMANCE
market via start-up companies; and some were incremental improvements to
existing products.
In the last 30 years, digital technologies have transformed the telephone
network from an analog system to a computer-controlled system with digital
switching and transmission. The process of digitalization has changed the indus-
try from two distinct businesses computers and communications to one busi-
ness in which computers and communications are intermingled in products and
services. This convergence was accelerated by advances in microelectronics and
increases in the bandwidth available for communications (Messerschmitt, 1996~.
The result is increasingly pervasive data networking, based largely on the packet-
switching technologies that emerged from academic and industrial research to
spawn the Internet.
The network systems and communications industry has a large and expand-
ing services component. For the telecommunications industry, which has always
been a service provider, the challenge is to invent and offer customers new,
valuable services that generate new sources of revenue. For the computer portion
of the industry, high-performance communications are making a wide range of
new services feasible. Examples include remote sensors and control systems;
integrated supply chain management systems; application service providers; full-
time, real-time stock quotes; and instant messaging.
DEFINITION OF THE INDUSTRY
The network systems and communications industry must be defined very
broadly. It clearly includes the manufacturing of telecommunications equipment
and the services that use such equipment, such as telephony, wireless telephony,
broadcast television, cable and satellite television, radio, and Internet service.
Both the equipment and services sectors increasingly require computing equip-
ment and software, and, in fact, the computer and communications industries are
no longer separate industries. For example, cellular telephony depends on a broad
range of technologies: the cell phone contains a liquid crystal display, an embed-
ded computer with a lot of software, and advanced chips that integrate most of the
components of a high-frequency radio; the transmission formats depend on ad-
vances in digital speech compression, signal modulation, and coding; the base
stations depend heavily on digital integrated circuits and computers for switching
and control and fiber-optic links between them; tracking a moving telephone
requires that computers at adjacent base stations exchange protocol messages for
the handoff; and the billing, provisioning, and maintenance of the service require
large-scale computing and software systems of the service provider. Separating
this integrated system into "communications" and "computing" components is
simply not possible. In short, computing and communications equipment and
services have converged, creating new business and technical opportunities.
OCR for page 31
NETWORK SYSTEMS AND COMMUNICATIONS INDUSTRY
31
The explosive growth of the Internet is the most visible manifestation of
this trend toward convergence. The technologies underlying the Internet-
just like those that underlie the cellular telephone include computing and
communications. Special computers serve as routers, and network services
knit together the transmission links and implement the collection of Internet
protocols that carry Internet traffic. The explosive growth of the Internet,
however, is attributable not to these basic provisions which existed before
1993 but to new services that created consumer demand: electronic mail.
the World Wide Web of information and its associated browser software; chat
groups; real-time delivery of audio and video media; online merchandising;
banking and financial transactions; supply-chain integration of suppliers and
customers; and numerous other applications. Some applications merely ex-
tend existing internal information technology systems to provide Internet
access. But others, such as eBay's success with online auctions, are entirely
new business concepts. As the Internet becomes more pervasive, old ways of
computing, in which data was created, stored, and manipulated at a single
site, are giving way to networked systems in which data can be accessed
remotely and shared extensively.
The computers embedded in everyday objects telephones, cars, televisions,
furnaces, hi-fi equipment are becoming increasingly capable and increasingly
networked. Some cars already can connect with a diagnostic and help center by
cellular telephone or satellite communications. Home networks in which multiple
personal computers in a household are linked over existing telephone wires and
short-range wireless devices will soon make networking of appliances routine. A
world in which all devices have an Internet address is not out of reach. Thus
computers increasingly require communications to fulfill their functions, and
communications increasingly require computers to fulfill theirs.
The technologies of computing and communications are becoming indis-
tinguishable. All of them depend on software to express functions at all levels
in the network. A few years ago, a modem was a complex, integrated circuit.
Today, with more complex algorithms and faster computers, modems are writ-
ten in software embedded within digital signal processors. Many algorithms
can be used equally well in computing and communications settings. For ex-
ample, schemes to digitize and compress video signals are useful both for
manipulating and storing video information on a computer disk and for trans-
mitting it over digital communications channels. Similarly, encryption technol-
ogy can be used to protect sensitive information in a computer system or in
transit over a network.
These three trends convergence, embedding, and network applications-
characterize the network systems and communications sector. The panel's assess-
ment of the contributions of academic research to this industry is based on this
broad definition.
OCR for page 32
32 THE IMPACT OF ACADEMIC RESEARCH ON INDUSTRIAL PERFORMANCE
TABLE 2-1 Sales and Employment in the Information Technology Industry, 2000
Sales Number
NAICSa Revenues of Jobs
Code ($ billions) (1,000)
IT Manufacturing
Computer and peripheral equipment 3341 $110.0 190
Communications equipment 3342 119.3 291
Software 5112 88.6 331
Semiconductors and
other electronic components 3344 168.5 621
IT Services
Data processing services 5142 42.9 296
Telecommunications services 5133 354.2 1,165
aNorth American Industrial Classification System. Source: U.S. Bureau of the Census, 2002.
Size
Because our definition has vague boundaries and because the industrial
classifications used to gather statistics have not been adapted to the rapid changes
in the industry, it is difficult to determine the size of the network systems and
communications sector. Table 2-1 summarizes sales and employment in the
information technology industry based on Bureau of the Census data (U.S. Bu-
reau of the Census, 2002~. Taken together, sales of computer and communica-
tions equipment and services (all information technology minus semiconduc-
tors) were about $715 billion in 2000, and the industry employed more than
2.2 million people (U.S. Bureau of the Census, 2002~. Expenditures for
information-processing equipment increased almost 10 percent per year on aver-
age from 1970 to 1994; the corresponding figure for computers and peripherals
was 27.5 percent (NRC, 1999~. A 1999 survey found that telecommunications
manufacturing was growing by 16.3 percent annually, computer software by
16.6 percent, and computer hardware by 9.5 percent (CTIS, 1999), however
these rates have dropped significantly since early 2000.
Structure
The role of research and innovation in the network systems and communications
sector can best be understood in the context of the structure of the industry, which
influences the mechanisms of innovation and thus how new technologies and prod-
ucts are introduced. The very general description that follows is intended only to
reveal similarities and differences with the other industries studied in this report.
Manufacturing
The structure of the computer industry is horizontal; the communications
industry was vertically integrated but has been rapidly changing to a horizontal
OCR for page 33
NETWORK SYSTEMS AND COMMUNICATIONS INDUSTRY
33
structure as well. In a horizontal structure, numerous suppliers manufacture parts
and components that many integrators assemble into subassemblies that are then
assembled into final products by numerous competing original equipment manu-
facturers. The multiplicity of companies at each manufacturing step ensures in-
tense competition throughout the production process, not only in terms of price
but also on a wide rage of performance characteristics. For example, manufac-
turers of personal computers buy disk drives from any one of about a dozen
suppliers. A company that needs a customized integrated circuit may design the
circuit but use one of several competing semiconductor foundries to manufacture
it. Specialized circuit board assembly firms can assemble and test complete cir-
cuit boards, giving an electronic design firm the ability to design and sell a unique
computer interface board with custom chips without having to invest in either
chip or board manufacturing facilities.
The divestiture of AT&T and the subsequent deregulation of communica-
tions services forced the communications industry to change from a vertical to a
horizontal structure. Today there are many vendors of telecommunications equip-
ment and components. Custom integrated circuits can implement very complex
communications functions; coupled with custom-built and proprietary software
designs, equipment vendors compete intensely in terms of technology, reliability,
and cost of ownership.
Another important feature of the network systems and communications sec-
tor is its reliance on components with well defined interfaces. Integrated circuits
are a good example: the physical, electrical, and logical behaviors of chip inter-
faces are specified by the manufacturer and used by the customer to determine
how to incorporate a chip into a subsystem with other components. Subsystems
then become components of still larger systems. Software, another component,
plugs into the operating system that supports it by linking the software interfaces
(sometimes called application programming interfaces, or APIs). A piece of soft-
ware that is compatible with a certain operating system adheres to the interfaces
provided by the operating system. Computer systems are built from complex
hierarchical assemblies of subsystems and components, sometimes hundreds or
even thousands of them. Some of the components are custom built, and some are
standard. Thus, interfaces give rise to components, which in turn give rise to
businesses structured around buying and selling components.
Key component interfaces become industry standards, which are usually
adopted by industry groups to hasten the spread of a new technology, increase
sales volume, and, therefore, decrease cost. Standards maintained by a group with
broad representation from competitive suppliers are said to be open standards.
For example, the Personal Computer Memory Card International Association is
an industry group that establishes standards for interchangeable interface cards
for laptop computers. The standards group includes several producers of cards
and several producers of laptops to ensure that the standard cannot be manipu-
lated to benefit one competitor over another. By contrast, standards promulgated
OCR for page 34
34 THE IMPACT OF ACADEMIC RESEARCH ON INDUSTRIAL PERFORMANCE
by a single vendor are said to be proprietary. For example, the programming
interface for Microsoft's Windows operating system is proprietary; Microsoft
specifies it and can change it at will.
Standards play a special role in communications. Broadly speaking, they are nec-
essary to ensure that components and subsystems connected via a communications
channel can operate together (e.g., they obey the same conventions for encoding voice
signals, multiplexing many simultaneous phone calls on a single channel, performing
operation and maintenance functions). Standards of this kind are necessary for guaran-
teed, sustained interoperability, and changes must be carefully designed to avoid even
slight interruptions of network service. New versions of standards must be designed so
they can be introduced incrementally, connect new equipment to old, test new proto-
cols, and so on. The same considerations apply to Internet protocols.
Services
Communication services (e.g., voice and data transmission, switching, and
distribution) are a major portion of the network systems and communications in-
dustry. The number and structure of telecommunications service providers have
been in constant flux since the divestiture of AT&T and the deregulation of local
telephone services. First, new companies emerged offering wireless telephone ser-
vices. Then another group of new companies emerged as Internet service providers.
To increase their revenue, carriers have been developing value-added services,
such as voice mail, call forwarding, call waiting, 800 service, electronic mail, and
virtual private networking, along with conventional transmission and switching
services. Internet service providers provide national and regional portals that offer
news, chat rooms, advertising, and direct access to the World Wide Web.
Computing services are also a major element of the industry. System integration,
the design and deployment of communications and information systems for large
clients, has become a major source of revenue for many equipment vendors. In recent
years, an important service has been to implement network capabilities across compa-
nies' existing computer systems. In some cases, networking has focused on providing
Internet access for employees and customers; in others, the focus has been on the
development of internal networks linking production and distribution facilities across
the company. So far, neither academic nor industrial research has addressed the
problems of service delivery in a structured and sustained manner.
INNOVATION SYSTEM
Most innovations are incremental improvements, such as design refinements,
improvements in technology and manufacturing processes, a better understanding of
customer needs, and integration of previously separate products. For example, impor-
tant performance metrics for communications equipment include low power and high
density (so that many circuits can be accommodated in the confined spaces of wiring
OCR for page 35
NETWORK SYSTEMS AND COMMUNICATIONS INDUSTRY
35
closets, boxes mounted on telephone poles, and even central switching offices). Both
power and density can be improved by advances in integrated-circuit technologies,
which in turn, derive primarily from incremental improvements in fabrication equip-
ment, processing steps, and materials. Research results may be the basis for some of
these improvements, and research has achieved major breakthroughs in these areas;
this research is performed or funded by materials, equipment, and microchip fabrica-
tors, not by the telecommunications equipment manufacturers (see Box 2-1~.
OCR for page 36
36 THE IMPACT OF ACADEMIC RESEARCH ON INDUSTRIAL PERFORMANCE
For many businesses, vendors of materials, products, and services throughout
the supply chain are major sources of innovation. Buying an integrated-circuit chip,
for example, implicitly buys a share of the dramatic improvements in price and
performance of integrated circuits (Moore's Law).i Over time, innovations will
make the chip faster or cheaper or more capable. A telecommunications carrier that
wants to deploy Synchronous Optical Network (SONET), a transmission protocol
that defines optical carrier levels and their electrically equivalent synchronous
transport signals, can purchase switches, multiplexers, and test equipment from the
vendors who developed SONET technology. This pattern is a direct consequence
of the "horizontal" structure of the industry. Dell Computer, for example, does not
have in-house R&D; in effect, Dell is a broker that negotiates attractive deals to buy
components and computer-assembly services for its build-a-computer-to-order busi-
ness. Dell depends on R&D investments by its vendors, especially Intel and
Microsoft, that make the microprocessors and operating system software on which
the personal computer business depends. Dell's innovations have been in its busi-
ness model and supply-chain management, not in its technology.
Innovation can also be purchased by acquiring other companies, especially
venture-capital-backed start-up companies that have introduced new products
with new technologies. A start-up company is a new business, often with an
innovative technology but with considerable risk. Often the innovative technol-
ogy has its origins in academic research. If the company makes good progress,
both in technology and in the market (e.g., beta testing, or success in getting its
approach adopted by standards consortia), it becomes an attractive target for a
larger company seeking to strengthen its technology or product line. For ex-
ample, Texas Instruments bought Amati; Fore Systems bought Berkeley Net-
works and Marconi bought Fore Systems; Cisco bought Granite Systems; and
Broadcom bought Epigram. Each of the acquired companies had ventured into a
new technical area. Epigram, for example, had devised a way to use home tele-
phone wiring to transmit 10Mb Ethernet traffic and had made progress in stan-
dardizing the scheme through the Home Phoneline Networking Alliance.
Broadcom, itself an innovative fabless chip company specializing in integrating
analog and digital functions of cable and twisted-wire modems, saw buying
Epigram as a natural way to enhance its core business.
Although high-tech start-ups seldom do research in the classic sense, many
behave much like "applied research" projects in an industrial laboratory. They
formulate technically aggressive plans based on established principles to pursue
and evaluate; the results of experiments often inform several products. For ex-
ample, Transwitch attempted to increase the telecommunications protocol process-
ing integrated on a single chip, as well as to partition the chip functions into an
"architecture" so that a small number of chip designs could be used to build a wide
variety of telecommunications products. Both Amati and Epigram conceived ways
of using advanced signal-processing techniques to adapt digital transmission to the
characteristics of real-world, twisted-pair copper wires (Amati) or in-house
OCR for page 37
NETWORK SYSTEMS AND COMMUNICATIONS INDUSTRY
37
telephone wiring (Epigram). The technology-development activities of these com-
panies are much like those in industrial research, but they are done in a commercial
setting and with strong incentives to bring innovations to market rapidly.
Industrial research is concentrated in the laboratories of a few of the largest
companies, such as Intel, Microsoft, IBM, Compaq, Lucent, AT&T, Hewlett-
Packard, Sun Microsystems, and Xerox. Although many of these firms invest 10 to
15 percent or more of revenues in R&D each year, the vast majority of this is for
"development," that is, for the engineering of the next generation of products.
Research focused on objectives more than 18 months or one or two product cycles
out is estimated to be, at most, 5 percent of that 10 to 15 percent, or far less than
1 percent of revenue.2 A few large companies eschew research, preferring instead
to buy innovative companies (e.g., Cisco). Companies in the services sector, how-
ever, generally do not engage in or support research. For example, at MCI, which is
generally considered a technology leader, the advanced technology group is primar-
ily concerned with testing new equipment and working with vendors to solve
interoperability and operation, administration, and management problems.
Industry research is usually driven by market needs but often includes some
fundamental or long-range projects as well. For example, IBM's research on the
Internet and electronic commerce includes some long-term work on cryptographic
systems for security and authentication. Industrial research often links advanced
technologies to emerging product needs. For example, as the Java programming
language became popular, industry laboratories at Sun, IBM, and elsewhere
launched projects to devise advanced techniques for the compilation, synchroni-
zation, and code simplification required for its implementation. Previous research
results in these areas had not adequately addressed the needs of the Java lan-
guage, of today's large memories, or of multiprocessor servers. Some of this
research is fundamental in the sense that it can be applied to problems other than
Java language implementations. In fact, even though research in engineering
fields is usually targeted toward meeting specific engineering needs, the results
are often useful for many other applications.
One of the companies' aims in operating research laboratories is to expand
their capability for bringing in new ideas and new people (Cohen and Levinthal,
1990~. The laboratory is expected to recruit people who cannot be recruited by an
engineering organization; it is also expected to interact with the intellectual
community by attending conferences, publishing papers, collaborating with uni-
versities, or entering partnerships with other companies; and it is intended to
counteract the risk inherent in the narrow focus of engineering projects on prod-
uct development.
A Culture of Innovation
Innovation in the network systems and communications industry can take
many paths. Even when research plays an essential role, there is no linear path
OCR for page 38
38 THE IMPACT OF ACADEMIC RESEARCH ON INDUSTRIAL PERFORMANCE
1965 1970 1975 1980 1985 1990 1995 2005
Timesharing
Client~server Commuting
e
;
Graphics
Interne1
LANs.
Workstations
Graphical user interfaces
VLSI design
I'
RISC processors
. , .
1965 1970 1975
| University ~ Industry R&D
CTSS, Multics / BED
Unix
SDS 940, 360/67, VMS
Berkeley, CMU, CERN
PARC, DEC, IBM
Novell, EMC, Sun, Oracle
Sketchpad, Utah
GM/IBM, Xerox, Microsoft
E&S, SGI, ATI, Adobe
Spacewar (MIT), Trek (Rochester)
Atari, Nintendo, SGI, Pixar
ARPANET, Aloha, Internet
Pup
DECnet, TCP/IP
Rings, Hubnet
Ethernet, Datakit, Autonet
LANs, switched Ethernet
Lisp machine, Stanford
Xerox Alto
(
to World Wide Web
1980 1985 1990 1995 2005
Xerox Star, Apollo, Sun
Engelbart / Rochester
Alto, Smalltalk
Star, Mac, Microsoft
Berkeley, Caltech, MOSIS
many
Berkeley, Stanford
IBM 801
SUN, SGI, IBM, HP
FIGURE 2-1 Examples of academic government-sponsored (and some industry-
sponsored) IT research and development in the creation of commercial products and
industries. Source: NRC, 2003.
OCR for page 39
NETWORK SYSTEMS AND COMMUNICATIONS INDUSTRY
1965 1970 1975 1980 1985 1990 1995 2005
Relational databases . ~
from/Tternet
Parallel databases
Datamining
Parallel computing
RAID/disk servers
Portable com inundation
I.... 1
~2
\--
World Wide Web
.
V,
. . ,
Speech recognition
Broadband in last mile
1965 1970 1975
\\ · ~ ~ ~ ~ -
1980 1985 1990 1995 2005
Berkeley, Wisconsin
IBM
Oracle, IBM, Sybase
Tokyo, Wisconsin, UCLA
IBM, ICE
JCL, Teradata, Tandem
Wisconsin, Stanford
IBM, Arbor
IRI, Arbor, Plato
Illiac 4, CMU, Caltech, HPC
IBM, Intel
CM-5, Teradata, C ray T3D
Berkeley
Striping/Datamesh, Petal
many
Berkeley, Purdue (COMA)
Linkabit, Hughes
Qualcomm
CERN, Illinois (Mosaic)
Alta Vista
Netscape, Yahoo, Google
CMU, SRI, MIT
Bell, IBM, Dragon
Dragon, IBM
Stanford, UCLA
Bellcore (Telcordia)
Amati, Alcatel, Broadcom
~ University
~ Industry R&D
· ~ ~ - - - Products
39
bit. market |
OCR for page 66
66 THE IMPACT OF ACADEMIC RESEARCH ON INDUSTRIAL PERFORMANCE
miniaturization slows, and the technical implications of these changes. Indeed, as
industrial research investments change, and as human capital stresses wax and
wane, it is important to keep long-term academic research activities alive, pre-
cisely because they are the long-lived seeds from which both ideas and people
can spring, regardless of the short-term financial health of the industry.
Recommendations
Recommendation 2-1. Universities and industry should take steps to ensure that
faculty and students are available to carry on research in computer science and
other information technology fields in the future.
Innovation, either from research or incremental engineering, depends on trained
researchers. Projected demand for computer science and other information technol-
ogy graduates indicates periodic shortages in coming years. To maintain the pipeline
of both academic and industrial researchers, the following measures could be taken:
.
Universities should Provide early research experiences for undergradu-
ates or even secondary school students.
Universities should provide career-development support for young fac-
ulty members.
· Fellowships should be provided for graduate students to encourage them
to pursue research degrees; industry should provide some of this support.
· Universities and industry should provide incentives for industry engineers
to return to academia for training in research.
· Universities should develop cooperative programs in which master's de-
grees are based not only on course work, but also on research experience.
· Training in academic research should include training in some of the
qualities students will need for jobs in industry.
Research should involve addressing not only small technical puzzles in
isolation, but also complex systems problems in context. Students should
be encouraged to confront complexity and to address real-world data and
operational problems.
· Research should encourage teamwork.
· High-caliber industry researchers and engineers should be encouraged to
take sabbaticals to work in academia, thus bringing real-world research
problems into academic settings.
Recommendation 2-2. Universities and industry should continue to develop
diverse collaborative arrangements.
OCR for page 67
NETWORK SYSTEMS AND COMMUNICATIONS INDUSTRY
67
Industry and universities should resist the temptation to impose standard struc-
tural mechanisms to promote collaboration. Incentives for personal interactions
between university and industry should be encouraged in the following ways:
· Provide support for strong, committed leaders and the collaborative orga-
nizations they lead.
Encourage sabbaticals in both directions, enabling academics to spend
time in industry, especially in start-up companies.
Support people and projects that involve academic and industry research-
ers in essential ways.
· Explore new ways to support personal interactions across academic-
industry boundaries, including using technology to support collaboration.
Recommendation 2-3. Universities and industry should make every effort to
invigorate academic research on networking.
The extraordinary success of the Internet and the lure of Internet-related
start-up companies have tended to focus attention on short-term goals, caus-
ing long-term research to suffer. The situation could be improved in the
following ways:
Acknowledge that the research community must take risks.
Focus academic research on the thorny problems of large systems: model-
ing, maintenance, upgrades, quality-of-service, security, and so on. Both
funding agencies and academics must recognize that large-scale systems
can best be addressed in a university setting. Even applied systems
research can be structured in a way that accommodates a long-
term approach.
Universities and funding agencies (and industry) should support long-
term, radical research on networks.
Universities and industry should encourage interdisciplinary research that
combines network technologies with design and social science disciplines.
Networked devices (especially hand-held mobile devices) will have to
meet both technical and human requirements.
Universities should recognize that valuable innovations and engineering
in the field are often not channeled through traditional peer-reviewed
publications. Therefore, effective industry interaction should be more
highly valued in decisions about academic promotion and tenure.
To revitalize academic research on networking, the National Science
Foundation should consider sponsoring a workshop on the subject that
brings together academic and industry participants. A new agenda could
provide a strong argument for industry support, either by individual firms
or by a consortium.
OCR for page 68
68 THE IMPACT OF ACADEMIC RESEARCH ON INDUSTRIAL PERFORMANCE
NOTES
1 Gordon Moore (cofounder of Intel) predicted in 1965 that the transistor density of semicon-
ductor chips would double roughly every year. See Moore, 1965.
2For example, during fiscal year 2001, Microsoft spent $4.38 billion on product research and
development activities excluding funding of joint venture activity. This represented 17.3 percent of
revenue that year. Microsoft Research, the part of the company that looks more than one or two
product cycles out, has around 600 employees and a budget of roughly $200 million, less than
5 percent of the $4.38 billion, or less than 0.8 percent of total revenue.
3The loss of faculty to commercial endeavors was limited in time and to only a few programs.
Data from the most recent Taulbee Survey of computer science and computer engineering depart-
ments indicate that faculty numbers have grown and are anticipated to grow through 2004. The
survey also indicates that faculty departures have ranged from 2.3 to 2.6 percent over the last several
years (Bryant and Vardi, 2002).
4Economists have long acknowledged "externalities," factors that alter the value of a good viewed
in isolation. Shapiro and Varian (1998) applied the idea to networks, so-called "network effects."
Robert Metcalfe, a popular speaker on the value of networks, has often said that the usefulness, or
utility, of a network equals the square of the number of users. This observation has been dubbed
"Metcalfe's law" (Gilder, 1993).
REFERENCES
AT&T and the International Computer Science Institute. 1998. AT&T Labs, ICSI establish Internet
research center. Press release October 8, 1998. Available online at: http://www.icir.org/
aciri.html. [June 24, 2003]
Bailey, D.E., F.S. Settles, and D. Sanrow. 1998. Designing, Controlling, and Improving SRC Re-
search Quality. Presentation at the NAE-Committee on Science, Engineering, and Public Policy
Workshop on the Role of Human Capital in Capitalizing on Research, Beckman Center, Irvine,
California, January 21, 1998.
BankBoston. 1997. MIT: The Impact of Innovation. Special Report of the BankBoston Economics
Department. Available online at: http://web.mit.edu/newsoffice/founders. [June 24, 2003]
Bell, D.G., D.G. Bobrow, O. Raiman, and M.H. Shirley. 1997. Dynamic Documents and Situated
Processes: Building on Local Knowledge in Field Service. Pp. 261-276 in Information and
Process Integration in Enterprises: Rethinking Documents, T. Wakayama, S. Kannapan, C.M.
Khoong, S. Navathe, and J. Yates, eds. Norwell, Mass.: Kluwer Academic Publishers.
Bryant, R.E., and M.Y. Vardi. 2002. 2000-2001 Taulbee Survey: Hope for More Balance in Supply
and Demand. Computing Research News 14(2): 4-11.
Brynjolfson, E. 1991. Information Technology and the "Productivity Paradox": What We Know and
What We Don't Know. Cambridge, Mass.: MIT Sloan School of Management.
Bureau of Labor Statistics. 2001. Table 3b. Fastest Growing Occupations, 2000-10. Available online
at: http://www.bls.gov/news.release/ecopro.tO6.htm. [June 24, 2003]
Card, S.K., T.P. Moran, and A. Newell. 1983. The Psychology of Human-Computer Interaction.
Hillsdale, N.J.: L. Erlbaum Assoc.
Cerf, V., and R. Kahn. 1974. A protocol for packet network intercommunication. IEEE Transactions
on Communications 22(5): 637-642.
Cohen, W.M., and D.A. Levinthal. 1990. Absorptive capacity: a new perspective on learning and
innovation. Administrative Science Quarterly 35(1): 128-152.
CTIA (Cellular Telecommunications and Internet Association). 2003. CTIA's Semi-Annual Wire-
less Industry Survey. Available online at: http://www.wow-com.com/pdf/CTIA_Survey_
Yearend_2002.pdf. [June 24, 2003]
OCR for page 69
NETWORK SYSTEMS AND COMMUNICATIONS INDUSTRY
69
CTIS (Corporate Technology Information Services, Inc.). 1999. Spotlight. Woburn, Mass.: Corpo-
rate Technology Information Services, Inc.
Cusumano, M. 1991. Japan's Software Factories: A Challenge to U.S. Management. Oxford, U.K.:
Oxford University Press.
de Sola Pool, I., ed. 1977. The Social Impact of the Telephone. Cambridge, Mass.: MIT Press.
Earl, M., ed. 1988. Information Management: The Strategic Dimension. Oxford, U.K.: Clarendon
Press.
Gilder, G. 1993. Metcalfe's Law and Legacy. Forbes ASAP, September 13. Available online at:
http://www.seas. upenn. edu/~gajl/metgg.html. [June 24, 2003]
Hill, S. 2001. Science and Engineering Degrees: 1966-1998. Arlington, Va.: Division of Science
Resources Studies, National Science Foundation.
Intel. 2003. Intel Research Network of Laboratories. Available online at: http://www.intel-research.net.
[June 24, 2003]
InternetWeek. 2001. Report: DSL Market to Rebound Next Year. October 24, 2001. Available online
at: http://www. internetweek.com/story/INW20011024S0005. [June 24, 2003]
Kahn, A. 1970. The Economics of Regulation, Vol. 1. New York: John Wiley & Sons.
Kahn, A. 1971. The Economics of Regulation, Vol. 2. New York: John Wiley & Sons.
Keen, P.G.W., and M.S. Scott Morton. 1978. Decision Support Systems: An Organizational Perspec-
tive. Reading, Mass.: Addison-Wesley.
Kraut, R., W. Scherlis, T. Mukhopa&yay, J. Manning, and S. Kiesler. 1996. The HomeNet field trial
of residential Internet services. Communications of the ACM 39(12): 55-63.
Lazowska, E. 1998. Remarks made during panel session on Changing the Interaction Between Aca-
demic Research and Industry: University, Industry, and Government Perspectives. Presented at
the workshop How Can Academic Research Best Contribute to Network Systems and Commu-
nications?, National Academy of Engineering, Washington, D.C., October 30, 1998.
Malone, T. 1998. Remarks made during panel session on Contributions and Impact of Academic
Research: Social, Management, and Policy Sciences. Presented at the workshop How Can
Academic Research Best Contribute to Network Systems and Communications?, National
Academy of Engineering, Washington, D.C., October 30, 1998.
Malone, T.W., K. Crowston, J. Lee, B. Pentland, C. Dellarocas, G. Wyner, J. Quimby, C.S. Osborn,
A. Bernstein, G. Herman, M. Klein, and E. O'Donnell. 1999. Tools for inventing organizations:
toward a handbook of organizational processes. Management Science 45(3): 425-443.
Markus, M.L. 1987. Toward a "critical mass" theory of interactive media: universal access, inter-
dependence, and diffusion. Communications Research 14: 491-511.
McKnight, L.W., and J.P. Bailey, eds. 1997. Internet Economics. Cambridge, Mass.: MIT Press.
Messerschmitt, D.G. 1996. The convergence of telecommunications and computing: what are the
implications today? IEEE Proceedings (August). Available online at: http://www.informatik.
tu-darmstadt.de/VS/Lehre/VES97-98/JavaTK/convergence.html. [June 24, 2003]
MICRO. 2002. University of California Microelectronics Innovation and Computer Research Oppor-
tunities. Available online at: http://www.ucop.edu/research/micro/. [June 24, 2003]
Moore, G.E. 1965. Cramming more components onto integrated circuits. Electronics 38(8): 114-
117. Available online at: http://www.cybraryn.com/news/documents/moore-paper.pdf. [June 24,
2003]
Morgan, R.P., and D.E. Strickland. 2000. U.S. university research contributions to industry: findings
and conjectures. Science and Public Policy 28(2): 113-121.
Morris, J. 1998. Remarks made during panel session on Changing the Interaction Between Academic
Research and Industry: University, Industry and Government Perspectives. Presented at the
workshop How Can Academic Research Best Contribute to Network Systems and Communica-
tions?, National Academy of Engineering, Washington, D.C., October 30, 1998
OCR for page 70
70 THE IMPACT OF ACADEMIC RESEARCH ON INDUSTRIAL PERFORMANCE
NRC (National Research Council). 1995. Evolving the High-Performance Computing and Commu-
nications Initiative to Support the Nation's Information Infrastructure. Washington, D.C.: Na-
tional Academy Press.
NRC. 1996. Cryptography's Role in Securing the Information Society. Washington, D.C.: National
Academy Press.
NRC. 1999. Funding a Revolution: Government Support for Computing Research. Washington,
D.C.: National Academy Press.
NRC. 2000. Making IT Better: Expanding Information Technology Research to Meet Society's
Needs. Washington, D.C.: National Academy Press.
NRC. 2003. Innovation and Information Technology. Washington, D.C.: National Academies Press.
NSF (National Science Foundation). 2001. Survey of Federal Funds for Research and Development:
Fiscal Years 1999, 2000, and 2001. Arlington, Va.: National Science Foundation.
Orr, J. 1990. Sharing Knowledge, Celebrating Identity: War Stories and Community Memory in a
Service Culture. Pp. 169-189 in Collective Remembering: Memory in Society, D.S. Middleton
and D. Edwards, eds. Beverly Hills, Calif.: Sage Publications.
Parker, L. 1997. The Engineering Research Centers Program: An Assessment of Benefits and Out-
comes. Arlington, Va.: National Science Foundation.
Qualcomm Corporation. 1999. ERICSSON and QUALCOMM Reach Global CDMA Resolution.
Press release. March 25, 1999. Available online at: http://www.qualcomm.com/press/pr/
releasesl999/press457.html. [June 24, 2003]
Rockart, J.F. 1981. The changing role of the information systems executive: a critical success factors
perspective. Sloan Management Review 22(2): 15-25.
Roessner, D., R. Carr, I. Feller, M. McGeary, and N. Newman. 1998. The Role of NSF's Support of
Engineering in Enabling Technological Innovation: Final Report-Phase II. Arlington, Va.: SRI
International. Available online at: http://www.sri.com/policy/stp/techin2/. [June 24, 2003]
Schofield, J.W. 1995. Computers and Classroom Culture. New York: Cambridge University Press.
Shapiro, C., and H. Varian. 1998. Information Rules: A Strategic Guide to the Network Economy.
Cambridge, Mass.: Harvard Business School Press.
Siegel, J.L., V. Dubrovsky, S. Kiesler, and T. McGuire. 1986. Group Processes in Computer-
Mediated Communication. Organizational Behavior and Human Decision Processes 37:
157-187.
Sirbu, M. 1998. Remarks made during panel session on Contributions and Impact of Academic
Research: Social, Management, and Policy Sciences. Presented at the workshop How Can
Academic Research Best Contribute to Network Systems and Communications?, National
Academy of Engineering, Washington, D.C., October 30, 1998.
Sproull, L., and S. Kiesler. 1991. Connections: New Ways of Working in the Networked Organiza-
tion. Cambridge, Mass.: MIT Press.
SRI International. 1997. The Role of NSF's Support of Engineering in Enabling Technological
Innovation. IV. The Internet. Available online at: http://www.sri.com/policy/stp/techin/
interl.html. [June 24, 2003]
Stanford University Corporate Guide. 2001. Top 10 Stanford Inventions. Available online at: http://
corporate.stanford.edu/innovations/invent.html. [June 24, 2003]
Tennenhouse, D. 1998. Diagram presented during panel session on Changing the Interaction Be-
tween Academic Research and Industry: University, Industry and Government Perspectives at
the workshop How Can Academic Research Best Contribute to Network Systems and Commu-
nications?, National Academy of Engineering, Washington, D.C., October 30, 1998.
Turkle, S. 1984. The Second Self: Computers and the Human Spirit. New York: Simon & Schuster.
U.S. Bureau of the Census. 2002. Statistical Abstract of the United States-2002. Washington, D.C.:
U.S. Government Printing Office. Available online at: http://www.census.gov/prod/www/
statistical-abstract-02.html. [June 24, 2003]
OCR for page 71
71
ADDENDUM
E-Mai} Questionnaire
The following questionnaire was sent to individuals selected from various
parts of the network systems and communication industry, some of whom at-
tended the October 1998 workshop. Included among the questionnaire respon-
dents were senior executives at AT&T Laboratories, Bell Atlantic, Bellcore,
MCI, and Motorola, and professors with expertise in computer science and engi-
neering, network systems, and telecommunications from Stanford University,
University of Delaware, University of California-Berkeley, University of Cali-
fornia, Los Angeles, University of Virginia, and University of Washington.
THE IMPACT OF ACADEMIC RESEARCH ON INDUSTRIAL
PERFORMANCE
NETWORK SYSTEMS AND COMMUNICATIONS PANEL
We invite your responses to the following questions. Your responses will be
used by our Panel as background information for our report. Any material used
verbatim will not be attributed to you without seeking your permission.
1. Could you describe briefly significant academic research contributions to
the network systems and communications industry? (If possible, please supply
references to published information that outlines the contributions.)
2. Overall, would you describe the impact of academic research on industrial
performance in the network systems and communications industry as (Please put
an X in one box):
1. very large
2. large
3. medium
4. small
~ 5. very small/non-existent
3. What is the role of academic research in educating people who work in
your industry? (Please focus on university research activities, rather than univer-
sity education generally.)
4. What structural forms of university-industry collaboration lead to good
results in your industry? An example of such a structure might be a discipline- or
industry-oriented "center" that solicits industry sponsors for a collection of
OCR for page 72
72 THE IMPACT OF ACADEMIC RESEARCH ON INDUSTRIAL PERFORMANCE
projects that span a varied research program. What seem to be the essential
determinants of success of such structures?
5. What are significant emerging trends or problems that the network sys-
tems and communications industry will face in the future that could benefit from
academic research?
6. What changes are required, if any, in academic research if it is to be
responsive to these industrial trends and problems?
7. What single step could be taken by universities to enhance the impact of
academic research on the industry?
8. What single step could be taken by companies to enhance the impact of
academic research on industry?
9. What single step could be taken by government to enhance the impact of
academic research on industry?
10. Do you see any downside to enhanced university-industry research col-
laboration? Things to be avoided?
11. Other comments? Any comments, pointers to other studies, or sugges-
tions would be appreciated.
OCR for page 73
NETWORK SYSTEMS AND COMMUNICATIONS INDUSTRY
WORKSHOP AGENDA
HOW CAN ACADEMIC RESEARCH BEST CONTRIBUTE TO
NETWORK SYSTEMS AND COMMUNICATIONS?
October 30, 1998
National Academies Building
2101 Constitution Avenue N.W.
Washington, D.C.
9:00 am Welcoming remarks and self-introductions
Wm. A. Wulf; President, National Academy of Engineering
73
9:15 am Overview of the work of the Network Systems and Communications
Panel and description of the wider NAE study
Bob Sproull, Panel Chair
10:00 am Break
10:15 am Session I. Contributions and impacts of academic research on
performance in the network systems and communications indus-
try: Engineering and the Physical Sciences
David Forney, Ambuj Goyal, Robert Kahn, H. T. Kung, David Mills
11:45 am Lunch in Meeting Room
12:30 pm Session II. Contributions and impacts of academic research on
performance in the network systems and communications indus-
try: Design, Social, Management, and Policy Sciences
Dan Atkins, Walter Bender, Robert Kraut, Tom Malone,
Marvin Sirbu
1:30 pm Session III. Structures for university-industry collaboration
James Flanagan, Stewart Personick, David Roessner,
Donald Strickland, Stephen Wolff
2:30 pm Break
OCR for page 74
74 THE IMPACT OF ACADEMIC RESEARCH ON INDUSTRIAL PERFORMANCE
2:45 pm Session IV. Changing the interaction between academic research
and industry: University, Industry, and Government Perspectives
Hamid Ahmadi, Ed Lazowska, James Morris, Rick Rashid,
George Strawn, David Tennenhouse
4:30 pm Discussion, conclusions and recommendations
Bob Sproull
OCR for page 75
NETWORK SYSTEMS AND COMMUNICATIONS INDUSTRY
WORKSHOP ATTENDEES
Robert Sproull, chair *
Vice President and Sun Fellow
Sun Microsystems, Inc.
Hamid Ahmadi
AT&T Labs
Alfred V. Aho *
Associate Research Vice President
Communications Sciences
Research Division
Lucent Technologies
Bell Labs Innovations
Daniel Atkins
School of Information
University of Michigan
Walter Bender
MIT Media Lab
John Cioffi *
Associate Professor
Department of Electrical Engineering
Stanford University
David J. Farber *
Alfred Fitter Moore Professor of
Telecommunications
University of Pennsylvania
James Flanagan
Center for Computer Aids
Rutgers University
G. David Forney, Jr.
Motorola, Inc.
*Parley member
75
Ambuj Goyal
IBM Corporation
T.J. Watson Research Center
George H. Heilmeier *
Chairman Emeritus
Bellcore
Robert Kahn
Corporation for National Research
Initiatives
Robert Kraut
Department of Social and Decision
Sciences
Carnegie Mellon University
H.T. Kung
Department of Electrical Engineering
and Computer Science
Harvard University
Ed Lazowska
Department of Computer Science and
Engineering
University of Washington
Tom Malone
Sloan School of Management
Massachusetts Institute of Technology
David Mills
Department of Electrical and
Computer Engineering
University of Delaware
OCR for page 76
76 THE IMPACT OF ACADEMIC RESEARCH ON INDUSTRIAL PERFORMANCE
James Morris
School of Computer Science
Carnegie Mellon University
Stewart Personick
Drexel University
Richard Rashid
Advanced Technical Research
Microsoft
J. David Roessner
School of Public Policy
Georgia Institute of Technology
Jerrard Sheehan
National Research Council
Marvin Sirbu
Information Networking Institute
Carnegie Mellon University
George Strawn
National Science Foundation
Donald E. Strickland
Chair, Management Department
Southern Illinois University
David Tennenhouse
Defense Advanced Research Projects
Agency
Stephen Wolff *
Executive Director
Advanced Internet Initiatives Division
Cisco Systems, Inc.
Wm. A. Wulf
President
National Academy of Engineering
NAE Program Office Staff
Tom Weimer, Director
Proctor Reid, Associate Director
Nathan Kahl, Project Assistant
Robert Morgan, NAB Fellow and Senior Analyst
*Parley member
Representative terms from entire chapter:
industrial performance
