Interesting People mailing list archives
report on Workshop on Basic Research in Telecommunication
From: Dave Farber <dave () farber net>
Date: Tue, 10 Jun 2003 17:12:07 -0400
Report Workshop on Basic Research in Telecommunication May 23, 2003 Sponsored by: Annenberg School for Communication at the University of Southern California Columbia Institute for Tele-Information at Columbia University Galileo Marconi International Fellowship Foundation at Columbia University. A. Michael Noll Rapporteur Contents SUMMARY..................................................................... .................................................................... 3 BACKGROUND.................................................................. .................................................................. 3 WHY IS BASIC RESEARCH IMPORTANT?.................................................................. ......................... 4 IS THERE A PROBLEM?.................................................................... ................................................. 5 DEFINING THE NEED FOR BASIC RESEARCH.................................................................... ................. 5 SUPPORT MECHANISMS.................................................................. .................................................. 6 Appendix A. Wall Charts...................................................................... ............................................. 8 Appendix B. Participant Comments.................................................................... ............................ 10 Appendix C. List of Workshop Participants:............................................................... ................... 12 Appendix D. Readings:................................................................... .................................................. 13 Report of Workshop on Basic Research in Telecommunication ³Leave the beaten track occasionally and dive into the woods. You will be certain to find something that you have never seen before.² Alexander Graham Bell (1847-1947) SUMMARY This report summarizes the deliberations of a workshop to discuss the current state of basic research related to telecommunication. A group of distinguished leaders of telecommunication research, most of whom have led or performed work at Bell Labs or one of its offspring, were invited to participate in the workshop, which was held on the afternoon of May 23, 2003 on the campus of Columbia University. The overall consensus was that there is a problem facing the future of basic research in telecommunication because of inadequate levels of support and increasing emphases on short term applications. These factors are largely a result of restructuring, the resulting competition, and the poor financial health of the telecommunication industry. The conclusion was that government support is warranted and required in the interest of innovation, economic growth, national security, balance of trade, and other national goals. While recommendations were made and potential models explored, the specifics of the mechanism for such support were unclear, as were the ways to assure relevance and connection between basic research and the practical problems of the telecommunication industry. The workshop raised more questions than it put to rest. That, in fact, was its purposeto help open the dialogue and catalyze a response, ultimately with a broader set of interested parties who are increasingly vocal in their concerns about basic research in the United States today. In capturing the workshop deliberations, this report therefore does not represent new or detailed research into the problem, but instead highlights the general views of the participants. BACKGROUND The fate of fundamental telecommunication research conducted at industrial laboratories seems to be at risk. Since the Bell breakup of 1984, this basic research previously performed exclusively at Bell Labs has become increasingly fragmented. Today the remaining research units are at AT&T, Telecordia, Lucent Technologies, Avaya Systems, and Agere Systems. Cutbacks and changing missions afflict these entities, thereby threatening the environment for basic research. The computer industry¹s labs are not filling the voidthey have only limited overlap in mission and face many of the same market constraints as the telecommunication industry. Academic research has not focused that strongly on telecommunication basic research and traditionally is challenged by the issues of mission and proximity to real problems. Yet, as it has in the past, the long-term future of telecommunication in the United States will continue to be shaped by advances in basic fundamental research. The decline in basic telecom research is the subject of increasing concern. On the very day the workshop took place, the front page of the Wall Street Journal heralded the stakes. Days before, the Bush Administration reconstituted the President¹s Information Technology Advisory Committee. In parallel, the National Research Council is leading a exploration into the subject. This fall, CITI will be holding an event to report on its policy recommendations for the recovery of the telecommunication industry. Despite concerns at these levels, however, the feeling among the general public is that there is a glut of technology90 percent, or more, of the potential capacity of U.S. fiber optic networks is currently said to go unused, and yet new high-tech advances in capacity continue to appear on the scene. WHY IS BASIC RESEARCH IMPORTANT? Usually, it takes from 10 to 20 years for the application of basic research to evolve into practical results. This long-term horizon has become difficult for industry to justify in today¹s competitive world with its emphasis on short-term profitability and incremental developments. Basic research is the unfettered exploration of new knowledge, resulting in discoveries and innovation. The conduct of the research is defined by the researchers themselves, rather than by concerns for improvements in current products and services. Knowledge and illuminating the way for the future are the goals of basic research. The history of the science and technology of telecommunication shows a steady stream of important discoveries and advances flowing from basic research, such as the transistor, the laser, microwave radio, and communication satellites, to mention just a few. This kind of innovation has contributed to affordable telecommunication for all people around the world and specifically to a telecommunication industry that generate nearly 3% of the gross national product of the United States. The new knowledge and discoveries that result from basic research create the foundation for surprises in the future of communication. The same knowledge also helps prevent unanticipated surprises. For example, electronic, computer-controlled switching changed telephone switching. Yet the basic concepts of electronic switching were discovered decades before its implementation in actual switching machines. There is a need for scouting by competent researchers of where technology is going, since this helps illuminate the future. Yet innovation appears to be on the decline in the United States. It is debatable whether this is unique to the United States or a worldwide trend. If unique, then the United States risks being left behind other countries in the discovery of new knowledge and its application to goods and services. Basic research is important to global competitiveness in another way. Developing nations seem able to manufacture anything more cheaply than anyone else. Thus with such globalism, the only way to stay ahead is to innovate. Those doing the basic research have clear ³first mover² advantages. There is a need for the real innovations that come from basic research in telecommunication. The future of telecommunication was determined very much by such research in the past. Today¹s world of telecommunication would be very sparse without these past innovations. These past discoveries and innovations indicate that research will result in future innovation, equally impressive and significant. National security is of much concern, along with issues of network reliability and vulnerability. A broader view of networks is required to supplement and even replace the disjointed views of individual carriers. IS THERE A PROBLEM? The overall majority of the participants believed strongly that there is a problem with the current state of telecommunication basic research. Some felt that the problems were specific only to certain areas. In comparing today¹s Bell legacy with Bell Labs before divestiture, it appears that the number of researchers has not changed. However, the numbers conceal what is really going on, namely, an emphasis on short-term practical applications rather than researcher-driven investigations. It was observed that ³horizons are getting shorter.² Today there are small numbers of self-directed people still doing basic research, but the numbers thus engaged are not sufficient for today¹s problems and for assuring the future. Also, much of today¹s research emphasizes incremental progress rather than the risky work that can produce real breakthroughs. What appears broken is the level of funding for basic research to create new ideas that emphasize longer-term horizons. Some of the participants felt that venture money was available for research. However, the long-term horizons of research with uncertain payoffs would most certainly not appeal to today¹s venture capitalists. The researchers are there and the real problems are there. The missing link in the chain is the lack of investment in supporting the research. DEFINING THE NEED FOR BASIC RESEARCH In the past, much research resulted in innovations in products that were ultimately manufactured and provided. Today¹s broader context is in services. The future challenge is no longer the physical layer that so stimulated much of the research of the past. Some research topic areas that were mentioned include: operations, networking, security and resiliency, database algorithms, provisioning, management of networks, worry-free networks, and network stability and interoperability. Most of these topics are service oriented in terms of their application. Transmission and switching enable connectivity. But today the provisioning of connectivity has become a commodity. Differentiation and value are created in the provisioning of services delivered over connectivity. This is a fundamental change in telecommunication with implications for research. A critical mass of researchers is needed. A few scattered researchers at different companies is not an effective way to perform the basic research that is needed for telecommunications. To justify the support of and the investment in basic research, a vision of the accomplishments and rationale for basic research in telecommunication needs to be articulated and promulgated. This vision must capture the imagination and demonstrate a national significance. SUPPORT MECHANISMS Businesses that are real or virtual monopolies (such as the old Bell System, IBM of the past, and today¹s Microsoft) are able to support their own basic research laboratories, since the expense can be subsidized by monopoly profits. Businesses in competitive industries are less able to do so, because the long-term horizons and broader relevance of such basic research are inconsistent with the near-term financial imperatives of competitive markets. Thus, governmentthe ultimate monopolyis really the only source of support for basic research to benefit competitive industrial sectors. Telecommunication today is much more competitive than in the days of the old Bell monopoly, and thus government support of basic research in telecommunication is now needed. Such service providers as the Baby Bells are not supporting any basic research. It seems difficult for them to make a rationale for such support. Instead, they rely on the research performed by the manufacturers, such as Lucent¹s Bell Labs. The Federal government should make the funds available for the support of basic research in telecommunication. This research should develop new ideas for the long-term future of telecommunication in the United States. The clear beneficiary of this research would be the country. Various sources of funding were discussed, such as tax credits, reprioritizations of existing funds, the 3% excise tax on telecommunications, spectrum auctions, and the Universal Service Fund. Political factors could make access to some of these sources quite challenging. There is a risk that tax credits and some sources of research funds simply disappear into companies¹ general budgets and do not directly support research. With the large amount of R&D funding by the Federal government, a small reprioritization could result in substantially increased funds for telecommunication basic research, which has been under funded historically, possibly because of the past extensive commitment by the old Bell System. Some form of endowment should be created over time to create financial security and independence. One possible source for such an endowment would be from the value of intellectual property. The old Bell Labs performed the function of a national laboratory in telecommunication basic research, producing a long list of discoveries and innovations with broad benefits to the United States and communications. Competitive pressures mitigate against industry support of such benevolence today. Thus government support of telecommunication basic research is warranted and needed to assure the long-term future of telecommunication and the role of the United States in guaranteeing innovation. There are many different venues for such basic research, including universities, university/industry partnerships, consortia, industry laboratories, and government laboratories. Each has their own advantages and disadvantages in terms of effectiveness, relevance, and excellence. U.S. government-funded research, in the past, has seeded such crucial developments as the Internet, and has advanced fiber optics (Multiwavelength Optical Network, or MONET). Both were sponsored by the Defense Advanced Research Projects Agency (DARPA). The National Science Foundation has provided important fellowships and funded research in the field. Priorities and processes within either of these organizations could be studied and recast, as necessary, to handle a larger portfolio of basic research in telecommunication. Some form of national telecommunication laboratory might be warranted. Such a national telecommunication laboratory should be a separate not-for-profit corporationnot a government laboratory. Some of the factors essential for success of such a facility include independence, competent management, links to relevant problems, a national rationale, and continuity of financial support. At its peak, before the divestiture of 1984, Bell Labs supported research at a level of about $300 million (1,200 researchers). Consideration should be given to achieving similar levels today through a combination of government and industry funding. Such funding should be considered within the context of the telecom industry¹s nearly 3% contribution to national GNP. Research needs freedom, but also exposure to real problems. Basic research must somehow be connected to the business models in the market. Lucent, to its credit, has attempted this, as expressed in its branding of ³Bell Labs Innovations.² The transfer of technology from research to application is essential. One possible way of transferring technology and assuring relevance is through the transfer and sharing of people. Industry people could come to the lab and visa versa. But it was observed that when this mechanism has been used usually junior people who are not essential are the ones sent, which loses any real impact. A consortium model for linkages to practical problems and technology transfer has been tried repeatedly but mostly with poor results. Bellcore is an example of a failed consortium to support telecommunication research. The management of the Argonne and Brookhaven laboratories could be possible models for a national telecommunication laboratory. So too could the Sandia Laboratory, which was managed for years by Western Electric, with close ties to practical problems. Fellowships, administered through the National Science Foundation, are an appropriate and successful mechanism to attract and support young researchers in telecommunication-related fields and topics. Appendix A. Wall Charts Terminology Basic Research Applied Research Systems Engineering Exploratory Development Development Number of Bell Family Researchers vs. Time Conditions for Research secure, stable, long-term funding sense of purpose & clear mission freedom to fail ties to real-world problems Research Fragmentation Discoveries & Innovations (representative list) RADIO ASTRONOMY HI-FI NEGATIVE FEEDBCK COMMUNICATION THEORY TRANSISTOR UNIX OPERATING SYSTEM CELLULAR WIRELESS LASER VOCODER ADAPTIVE ECHO CANCELLER COMPUTER ART & MUSIC SOLID-STATE PHYSICS PULSE CODE MODULATION FAST FOURIER TRANSFORM ELECTRET MICROPHONE PHOTOVOLTAIC SOLAR CELL ERBIUM-DOPED FIBER AMPLIFIER CCD IMAGER COMMUNICATION SATELLITES Appendix B. Participant Comments The participants in the workshop were given an opportunity to comment on a draft of this report and to express their own thoughts. This section presents those comments. Rob Calderbank When there was one Bell System, there was vertical integration of service infrastructure, network / service / customer management, and network infrastructure within a telecommunications monopoly. The business model of guaranteed return on investment made it possible to use service revenue to fund basic research across a broad spectrum of research disciplines. In fact this model encouraged unfettered exploration, since the greater the investment, the greater was the return. Vertical integration is no more, and today, the telecommunications industry is fragmented into different horizontal layers. Service infrastructure supports the business solutions practice that is of primary importance to IBM and others, network / service/ customer management is the core competence of multiple carriers as well as an emerging service industry, and Lucent is one of very many network infrastructure providers. These different layers have very different economics. Today connectivity is a commodity business, and success is tied to sales to the enterprise, and to sales at the edge of the network rather than the core. The higher margins are associated with value added services, for example, running distributed applications across wide area networks. What IP networking does is reinforce this separation of service value from network connectivity. It may be that connectivity became a commodity business through the madness of over investment in telecommunications, and that margins may return over time, but commodity businesses can not be expected to sustain significant long term investment in basic research. Today, all network infrastructure players are embracing network and service management as a new growth business, and are reorienting basic research to support the shift. There is not a strong tradition of support for basic research within telecommunications carriers, and where there are long term investments they are tightly coupled to improving operational efficiency, so that even where there is basic research, it is a means rather than an end. On the other hand, there is a long tradition of support for basic research within IBM and HP, and a shorter history at Microsoft. Within IBM, I think it would be instructive to look at total numbers over time, but particularly at the distribution between IT infrastructure and services. This analysis is consistent with the observation of declining support for basic research within the Bell family of companies and increasing emphasis on custom network solutions. There is in fact reason to consider the near term value of innovation to carriers and network infrastructure providers that are encumbered by significant legacy businesses. For example, advancing the practicability of VoIP cannibalizes legacy voice revenue, and direct wireless connection to an IP network substitutes for legacy cellular connection. There has been more innovation at the network edge and this is consistent with there being less friction associated with introducing innovation in services. Federal support for a national telecommunications laboratory can only be justified by importance to the nation. Even if we were to find that the Bell family of companies has fallen short in its support for fundamental research, it would be necessary to identify consequences in terms of global competitiveness, or national security that are sufficiently compelling to justify federal funding. Certainly there is reason to be concerned about access to broadband, the stability of IP networks, and the security of IP services. A. Michael Noll In order to justify the support of and the investment in basic research, a vision of the accomplishments and rationale for basic research in telecommunication needs to be articulated and promulgated. This vision must capture the imagination and demonstrate a national significance. Of basic research in telecommunication. This might imply a need for public illumination and education of the role of telecommunication basic research. Some form of telecommunication museum (perhaps at the Smithsonian or at some location in New Jersey where so much of the research was performed) might make sense to help achieve this goal. Another mechanism would be a focused media campaign, using appropriate researchers to humanize the message and also to attract young people to this field. Clearly attention should be given to these and other media relations and public relations aspects of telecommunication basic research. What seems required is a spokesperson who can clearly articulate the need and mission of telecommunication basic research, drawing on the relevance of past accomplishments and expressing excitement for the future. In the past, John R. Pierce and William O. Baker of Bell Labs very much had that role. The computer industry has quite a few very visible advocates, telecommunication research needs equivalent spokespersons and advocates. A ³national telecommunication laboratory² might seem warranted, managed in a corporate style but funded by the government. A number of smaller laboratories, located in a variety of venues, but linked together might also function as such a national resource in a manner loosely similar to the old ARPA community. The key conclusions, though, seem to be that basic research in telecommunication has suffered as a result of the fragmentation of the telecommunication industry, that such research will continue to be essential to the future of telecommunication, and that some broader form of support is required. The precipitous drop in the number of researchers seems to be a trend that will continue. But numbers alone do not tell the story. There has been a radical change in the nature of the work that is classified as ³research,² with the majority of it today being much more of an exploratory and applied nature. While in the past of a few decades ago, over 90% of the work classified as research was fundamental basic research, today the feeling is that only 10% (or even less) is basic research. Most of ³research² today is incremental in nature or in support of existing products and services. Appendix C. List of Workshop Participants: Industry, Government, Academic Representatives: Robert Calderbank AT&T Labs David Farber University of Pennsylvania A. G. (Sandy) Fraser Fraser Labs Jeffrey Jaffe Bell Labs Research John C. B. LeGates Harvard University Robert W. Lucky Retired, Telcordia Andrew M. Odlyzko University of Minnesota Mark Pinto Agere Systems Lawrence R. Rabiner Rutgers University Casimir Skrzypzak Retired, NYNEX Edmond J. Thomas Federal Communications Commission Sponsor Representatives: Robert C. Atkinson Columbia Institute for Tele-Information Darcy Gerbarg Marconi Fellowship Foundation John Jay Iselin Marconi Fellowship Foundation Eli Noam Columbia Institute for Tele-Information Chair: Karen Lynch Markle Foundation Organizer & Rapporteur: A. Michael Noll Annenberg School at USC Sponsors: The workshop was sponsored jointly by: the Annenberg School for Communication at the University of Southern California; the Columbia Institute for Tele-Information at Columbia University; and the Galileo Marconi International Fellowship Foundation at Columbia University. Planning and arrangements were performed by the Marconi Fellowship Foundation, and the Annenberg School provided financial support. The workshop was coordinated with a panel of the National Research Council that is studying research in telecommunication and was held with the cooperation of the Markle Foundation. Appendix D. Readings: Noll, A. Michael, ³Bell System R&D Activities,² Telecommunications Policy, Vol. 11, No. 2 (June 1987), pp. 161-178. Pierce, John R., ³Some Thoughts About Laboratories And Research Therein,² June 13, 1986, unpublished piece. Noll, A. Michael, ³The Effects of Divestiture on Telecommunications Research,² Journal of Communication, Vol. 37, No. 1 (Winter 1987), pp. 73-80. Pierce, John R. & A. Michael Noll, SIGNALS: The Science of Telecommunication, Scientific American Books (New York, NY), 1990, ³The System That Was,² pp. 219-223. Noll, A. Michael, ³The Future of AT&T Bell Labs and Telecommunications Research,² Telecommunications Policy, Vol. 15, No. 2 (April 1991), pp. 101-105. David, Jr., Edward E., ³Science in the Post-Cold War Era,² The Bridge, Spring 1994, pp. 3-8. Odlyzko, Andrew M., ³The Decline of Unfettered Research,² October 4, 1995, unpublished essay. Noll, A. Michael, ³The communications revolution began in New Jersey; it may end here,² The Sunday Star-Ledger, April 21, 1996, Section Ten, p.5. Stix, Gary, ³The Relentless Storm,² Scientific American, March 2003, pp. 42-45. Noll, A. Michael, ³Telecommunication Basic Research: An Uncertain Future for the Bell Legacy,² Prometheus. 2003. Accepted for publication. ------------------------------------- You are subscribed as interesting-people () lists elistx com To manage your subscription, go to http://v2.listbox.com/member/?listname=ip Archives at: http://www.interesting-people.org/archives/interesting-people/
Current thread:
- report on Workshop on Basic Research in Telecommunication Dave Farber (Jun 10)