Click for next page ( 30


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © 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 29
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 29
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 29
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 29
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 29
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 29
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 29
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 29
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 29
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 29
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 29
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 29
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 29
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 29
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 29
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 29
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 29
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 29
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 29
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 29
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 29
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