We live in an era of change in modes of communication. Few countries on earth have not already been effected by electronic broadcast technology. Today more and more nations are feeling the increasing effects of modern two-way electronic communication in the forms of cellular telephones and fax machines.

Today's emerging communications media differ from those of the past in a singularly im-portant way. Those that had dominated the majority of the twentieth century did not inter-operate; telephony, telegraphy and broadcast media stood alone and apart from each other. But, during the 1980's each of these modes went digital, in the wake of computing technology. As a conse-quence, notwithstanding their perceived differences, voice, music, text, graphics, motion video, numerical data and computer programs have all entered the domain of digital electronics. Today each can be conveyed across a common digital network.

With the hope of a National Research and Education in America, a new emphasis is being placed on converging these modes of communication, now in digital form, with inter-institutional networking. Thus, the scene is set for brand new technological approaches to instruction, learn-ing, and research.

This paper deals with one important technological component in what could become a global infrastructure for conveyance of digitized information. It has to do with wireless communication of digital data over very large areas through a series of protocols that enable signals to move end-to-end in a highly populated invisible network. This technology is called packet radio.

In a nut shell, packet radio shares many of the characteristics of the ETHERNET or Token Ring protocols, but without the cable and the short distance limitations. Like radio technology on which it is partially based, it is wireless. Packet Radio networks require no physical right of way. Packet radio networks can operate at fast or slow speeds, and they can be designed to achieve extremely efficient use of the electromagnet spectrum, in contrast to earlier twentieth century methods of radio. Like packet switching technology on which they are partially based, multiple packet radio networks can be bridged and gatewayed.

While the history of wireless communication itself is relatively short compared to the his-tories of other modes of communication, the history of combined wireless and packet switching technologies is briefer still. To date, the predominant research and development of packet radio systems has been by amateur radio operators and by the military establishment.

In developed countries it seems unlikely that packet radio systems would grow quickly, if at all, as a means of common carriage, because the digital communications market is already crowd-ed by regulated common carriers who enjoy real demand from large customer bases and who already have very large investments in other emerging bulk technologies such as fiber optics. In addition, in developed countries, the electromagnetic spectrum is regulated in extreme, and making rules for packet radio in common or private carriage would require a very slow process of making or changing rules.

Developing countries, on the other hand, are in a good position to exploit the potential of packet radio for high-speed wide area networking within and beyond their boarders. The theory and practice of packet radio technology is sufficiently well understood that it is feasible to introduce high speed, flexible, reliable metropolitan and wide area networks using packet radio technology

in conjunction with satellite technology.

2. BACKGROUND

My interest in packet radio originally arose while I was at the University of California as Director of Library Automation. It was originally sparked in 1981 by the development of a special purpose microprocessor from the Tucson Amateur Packet Radio (TAPR) Corporation. The TAPR board pioneered low-cost interfacing between a personal computer and a radio transceiver. It made it possible to move data packets across continents through a series of intervening packet radios. The service was relatively fast, reliable, and inexpensive, compared to that available from American common carriers.

Later, supported by grants from IBM, the Council on Library Resources, the U.S. Depart-ment of Education and the California State Library, we built several packet radios that used frequency shift keying and ran at a rate of 200 KB per second. One exotic experiment placed the 200 KB signal inside of a television transmission, between the suppressed chromonance carrier and the high volume threshold for the FM audio. There was no mutual interference, as that portion of the spectrum within a National Television Standards Committee (NTSC) signal is otherwise not excited. But, while we enjoyed magnificent cooperation from the U.S. Federal Communications Commission (FCC), it became obvious that without a rules-change process requiring millions of dollars in legal fees over a period of several years, our activities would remain strictly experimental.

Toward the end of our experiments at the University of California, we concluded that in American libraries represented only one possible user community for high speed fixed packet radio networks. Others included hospitals and health professionals outside of urban areas, faculty at educational institutions, in fact, any practitioner or researcher requiring digital networking services.

But our final conclusion was that in America the regulatory environment for non-military, non-amateur packet radio was not good. This follows from the FCC's traditional view that the spectrum is a very limited public good, and that its regulators must make choices about spectrum use among competing licensees.

The case may be very different in developing counties where often the common carrier infra-structure is based on wire, cable, fiber and point-to-point microwave technology, including satellite use, but where there also is relatively greater unused spectrum. In fact, due to the topology of much of Latin America, for example, radio technology is the only feasible means for communi-cations.

Because of satellite technology many developing counties have become communications rich over night, but with the ironic consequence that often it is more easy to communicate with another country that within one's own. Within the countries boarders, common carrier service may still be analog-based with long lead times for new services.

In developing countries where little, if any, digital communications infrastructure exists, packet radio technology may be the solution in search of a problem. Packet radio systems have been designed that could operate at speeds on the order of 10 MB and with proper network management they could allow international standard services such as ISDN. In combination with satellite systems, regional packet radio networks could span large geographical interstices.

Not only is this true for rural populations, but business centers in densely occupied down-town areas could benefit from packet radio "switches" installed at customers' premises. One has only to view the wires strung among office buildings in business districts of large cities in develop-ing countries to appreciate the obviousness of the application.

3. CONVERGENCE

Remembering that all modes of communication are digitizeable, and that given public policies that translate to spectrum availability, packet radio communication could play an important role in what we at the Memex Research Institute call the "personal network." In 1945 Vanavar Bush had a vision of the Memex, a device for your personal memory. If Bush had based its memory on magnetic media instead of photographic media, he aptly would have been foreshadowing the per-sonal computer.

Today, the personal computer is becoming ever more common in the educational environ-ment. It certainly has its pragmatic use in word processing, spread sheet work, and CD-ROM data base access. Much of the hype of computer-based learning has not been realized, because the hardware base continues to advance so as to obsolete expensive courseware development. This is in contrast to the humble successes of instructional television, the staying power of which has been guaranteed by the fact that the medium of TV is rigorously standardized and thus not subject to volatile changes such as those seen in the computing industry.

Contrasting computer-based learning to traditional instructional television suggests two things. Firstly, to avoid the disappointments of courseware obsolescence, a fundamental principle in usefully harnessing the computer in the learning process is to stay within the bounds of com-puting's underlying standards. Secondly, as an educational medium computers are potentially much more powerful than the analog television medium, because computers are capable of multi-way communication, whereas television is inherently a one-to-many, one-way medium.

In fact, with the advent of digitally compressed video, full exploitation of the networking capabilities of personal computers can be realized for applications never dreamed of by Vanavar Bush. With a personal computer as modest as the old IBM PC AT configured with a VGA monitor and a 128 KB network connection, learning takes on new dimensions.

Looking back on the 1970's and 1980's we saw the explosive growth of the Internet. By 1990 over sixty online library catalogs were accessible across the Internet. Today there is serious planning for adding databases of still images as well as moving images with sound.

The still images first will come about as book and journal material is scanned and stored. Not only will this be an alternative to microfilming material for the sake of preservation, but it also will be a by product of CCITT Group IV-based inter library loans. The moving images with sound will follow logically as there is a growing demand to access the recording of our own history as captured by news media and interpreted by cinematographers and others.

Non-scholarly applications such as surrogate travel, video conferencing, games and other activities we have not even thought of will become part of this new interactive world.

In developed countries this is likely to take longer than in developing countries, because in developed countries the telecommunications infrastructure is not sufficiently flexible to handle all digital communications at 128 KB or higher. Moreover, since, as mentioned above, there is a perceived scarcity of spectrum, in developed countries wireless communications does not have the potential to leap-frog the old infrastructure.

In developing countries where spectrum is more plentiful, there is an opportunity to create wholly new markets based on wireless packet switching. It may be that high priority human activity such as in the medical field will be the trigger application for networked interactive video. Or, maybe networked interactive video will become a new medium for advertising and sales. In any event, there is a totally new opportunity for instruction, education and research, at the very least.

For years lofty idealism about universities without walls has abounded. There are even those who talk about "universities of the world." Without a supporting communications infrastructure, these remain merely visions. But, with an infrastructure such as the one described here, such visions can materialize.

In the 1980's desk-top computing became a new market and had a noticeable impact on education in America. To a limited extent slow speed networking via telephone dial-up also had an impact on library-related services, and to a large extent high speed networking via the Internet gave America a competitive shot at supercomputing. Within a decade personal computing began to tend toward personal networking, but only for a few, only for those connected to the Internet. The real potential for personal networking is only now becoming obvious.

Where might personal networking start to become the next logical extension of personal computing? It will not be in America, nor in Europe, but rather in the Pacific basin, in Latin America, in China, in Africa, in the Soviet Union, where wireless packet switching can be deployed without instituting controversial public policies.

4. REGULATORY ISSUES

Assuming real user demand, and assuming the availability of electromagnetic spectrum, then the real public policy issue about packet radio networking is how to regulate it as a public good in the public interest. Should it be run by the national government, such as radio and television broadcasting in Europe have in part been? Should it be a regulated commercial activity like broad-casting in America? Should it be regulated like common carriage in America?

Broadly speaking, there are no simple answers to these questions. Most countries have at least two criteria in common when it comes to regulating the air waves, namely, who is using the air waves and for what purpose. Perhaps more than one regulatory policy would be warranted depending on who is using the air waves and for what. A body of regulation for advertising and selling via personal networking with packet radio might look very different for personal networking in support of higher education.

In the case of higher education, the Internet in America represents something of a public policy paradox. True enough, the circuits on which the network runs are subject to common carrier regulation. But, no regulatory policies at all are applied either as to what goes over the network or as to how its use is paid for.

Further scrutiny of the Internet in America reveals that in addition to the necessity of having a circuit that connects one into the Internet, and therefore a willing partner at the other end, there is only one single rule to be followed in order to partake of services on the Internet -- each computer on the internet must be running Transmission Control Protocol/Internet Protocol (TCP/IP). All other practices on the Internet are voluntary and by convention.

After more than twenty years of operation, the Internet community is huge. No one really knows how many computers are connected to it, but they must number in the hundreds of thou-sands. That may translate into over a million people in America whose lives are affected by what they do with the Internet. The paradox is that these people are engaged in personal networking, and no one is regulating, or collecting a fee for, their activity. The nearest thing to management within the network is at the level of regional groups who have volunteered to work together in order that traffic across the whole Internet is properly routed and load leveled.

The backbone circuits of the Internet are funded by the National Science Foundation, but connections among Internet nodes are paid for as part of institutional overhead. No effort is made to recharge use back to the people engaged in personal networking.

The U.S. Congress is currently debating a bill that, if passed, would create the National Research and Education Network (NREN). The existing Internet would comprise the underlying technology, but what rules should be applied to the NREN is still subject to debate. In a paper that I prepared for the Library Information and Technology Association, I proposed adoption of several principles, originally developed by Ithia de Sola Pool, as a basis for the public policies that would apply to the NREN. (Edwin Brownrigg. Developing the Information Superhighway: Issues for Libraries. Memex Research Institute, 1990.) Essentially these proposed principles are a combination of First Amendment Rights of expression and laissez-faire economics.

If these principles were applied to personal networking via packet radio in developing countries the result would be networks that looked and behaved much like the Internet in America today, except that there would be no common carrier circuits, and the price of admission would be a personal computer and a packet radio, of which the former is the more expensive.