Directions for the information highway.

You won't find a map for the information highway yet, but the road signs are going up fast. If you want to stay ahead of the traffic, read on to get your bearings.

For centuries, the seven wonders of the world represented the zenith of human civilization. Today, the computer chip and our communications network are two of the great wonders of the modern world -- but, unlike the original seven, these aren't fixed in stone. As a matter of fact, we're now completely re-inventing the network we've spent the last century building. Translation: chaos.

What are some of the evolutionary forces emerging from this chaos and driving the network's development? Fiber optics is certainly one of the most powerful technologies we have. In just the last few years, it's increased our telecommunications capacity enormously. A single fiber today carries 2.5 billion bits per second. And the technological capacity to send more per fiber is doubling every year and will continue to do so for the next 15 years.

As we begin to digitize our communications network, the first problem is getting this capacity into homes and businesses. Getting it to large businesses is no problem because they buy their own fiber and put it in themselves. Small businesses and homes are another matter.

All told, we'll probably have to spend about $23 billion to fiberize the nation. That's a large investment, but not impossibly large. At the current rate of investment and production, fiberizing would take about 15 years, but it'll probably happen faster than that. What's more, we may end up wiring the nation's homes at least two and possibly three times. You don't need two fibers going into your home, but because the market, not a regulatory body, is running the show, everyone wants to get into the act -- the cable television industry, the telephone people and the entrepreneurs, not to mention the electric power utilities, which see a chance to get into the network business themselves.

So, one way or another, optical fibers will come to the home. Many different kinds of architectures are possible, but right now none of them has compelling advantages economically or technologically over another. That's why this process will develop gradually over the next six to 10 years, and the market will expend a great deal of energy to make sure it happens.

Of course, when you're talking about fiberizing the homes of America, the money's in entertainment. There's no argument about that, so the cable television and telephone industries will be gearing up to provide an entertainment infrastructure. In fact, Vice President Gore and his advisers see video for the home as one of the principal components of the national information infrastructure.

Another great theme in telecommunications is wireless mobility -- sort of the unwiring of America. This is another way of getting information into the home, and it's playing out much faster than anticipated. For example, we're quickly moving toward personal-communication systems. The idea is to have many small bay stations hanging on the sides of buildings and on telephone poles, so you're never more than 500 feet from a station. As you walk along, your wristwatch telephone talks to the nearest bay station. In essence, you have a personal-communications network, possibly centered around a mobile telephone number that follows you from place to place. Right now this concept doesn't have the broad-band capabilities of optical fibers, so you can't transmit certain materials, like video. But we may eventually have the technology to do that.

At first glance, wireless mobility may seem incompatible with fiber optics. But neither will displace the other. Different economics and capabilities come into play with each technology, so there's no clear winner, at least not yet.

If fiber optics and wireless mobility are the drivers of the communications network, asynchronous transfer mode and Internet are the vehicles. ATM is a technology that packages data into 53-byte envelopes (48 bytes with a five-byte header), and its development and applications are unfolding at a fantastic rate. Suddenly, ATM is taking the telecommunications world by storm, because the computer industry and applications developers bought into it early and worked together to create an entirely new paradigm for communications. ATM makes the old paradigm look like two tin cans and a string.

With ATM, you don't have a physical connection to the network. It's more like a virtual post office. You drop "letters" (your files) into it and they're routed around, and each letter is independent of all the others. The letter carries an implicit address because the connection paths or virtual channels are set up to deliver it to the right place.

ATM technology has many virtues for local-area networks. ATM LANs, which are just coming out, have a very high bandwidth. They don't rely on a bus or shared-wire structure like Ethernet and other LAN buses. Instead, the structure depends on the user population. Every person has a dedicated line of 155 megabits. For your corporation, that means ATMs have the potential to expand band-widths for distributed computing.

And as the ATM infrastructure penetrates the network, you'll eventually be able to make inter-enterprise and long-distance links locally with the help of your ATM LANs. Distributed computing will enable you to have highly efficient, dispersed operations. You can prepare your business for these developments by gradually converting your LANs to ATMs. You'll have to wait a few years before you can do long-distance networking, but you can already do it in the local environment. By the end of 1994, the market could start taking off.

Just as optical fibers have made transmission cheap, ATM technology will make switching cheap. ATM switches actually have more power than the giant switching offices in the telephone plant. They don't do exactly the same thing, so it's not a completely fair comparison. But for $20,000 to $50,000, you can get an ATM switch that interfaces to the LAN and routes information packets to the network -- a huge amount of capacity.

GET SET FOR INTERNET

That large potential also holds true for Internet, one of the definitive models for the national information infrastructure. Internet's been around for more than 20 years, and for a long time no one took it seriously because it was a network for academics. Even as recently as three years ago, most executives I was working with at AT&T had never heard of Internet. But in the last three years, this has changed dramatically. Internet is something you should watch and take advantage of in some way. Many corporations have very efficient, powerful corporate networks, but what they don't have is access to the public networks connecting them to the information marketplace. Internet is providing the infrastructure to allow that to happen.

Internet's like the Esperanto of computer networks, and it's growing at 100 percent per year. Contrast that with the 3 percent to 5 percent growth in the telecommunications network overall. Internet connects about 20 million users and 22,000 networks.

The National Science Foundation operates Internet's national backbone. Not surprisingly, many business people believe that Internet's government subsidies are one of its disadvantages. That's irrelevant. If all the government subsidies were removed, the per-user cost would increase by only 20 cents per month. So don't think of it as a subsidized network. It's essentially commercial -- in fact, it now exceeds 50-percent commercial use. In the last few years, the free national networks have been augmented by commercial information-exchange networks that carry pay traffic across the country in the same Internet protocol format, so that we now have a pay path and a free path.

Regional networks are the new catalysts for Internet use. Often, they're little ma-and-pa companies that spring from computer centers in universities, where they buy some routers and go into business as a regional network. You can buy a private line into one of these networks, and it'll connect you to the national network. Almost all the costs you incur are for administering your local information environments. The costs of connecting to the regional and national networks don't add up to much -- about $20,000 to $30,000 a year for a corporation. Many executives are willing to foot the bill for these expenses. They reason that they'd incur these costs anyway just to run their computers, so Internet is really a free add-on.

Personally, I agree with that philosophy. I pay $35,000 a year for the routers and support for the Internet connection, $7,000 a year to lease a private line from New Jersey Bell into JVNC Net and $37,000 a year to connect to Internet. For the paltry sum of $80,000 I get all the communication 3,000 people need, without any usage charges. At approximately $27 per person, it's an incredible bargain, even if it does violate the rules of life and business -- that someone must be in charge and someone must pay.

Right now Internet doesn't support voice and video very well, but a new protocol is in the works that will support high-quality speech and video, so you'll be able to place voice telephone calls free through the network. This will probably break the traditional telephone archetype, so it will be an interesting development to watch.

Plus, a tremendous market is opening up for anything that would bring Internet into broader public use. Last year, the supercomputing application center at the University of Illinois came out with an information navigator called Mosaic that's been a phenomenal success. In only nine months it grew from zero to close to two million users, and it's still growing at 11 percent per week. Mosaic takes over the mechanics of Internet for you and connects you with distributed information sources around the world, creating an information marketplace in which you can put information onto Internet and other people can either receive or buy information from you. And you can see pictures and get movies through Mosaic, without knowing anything about how the program works.

FAQs is another great Internet invention. It stands for frequently asked questions. It's a great way of finding information because it's like going to a reference library and getting all the information it has on a given topic. For each question, Internet has a keeper somewhere who's compiled all the questions people ask about the topic and who's amassed all the expert answers. You can download that file and find out all kinds of really useful, practical information about your topic.

Internet has plenty of other creative uses. Several art museums are on line, not to mention the Library of Congress. The pictures from the space probes are all on line, and so are the Vatican archives. Soon you'll be able to get sound movies.

THE EYES HAVE IT

Once ATM proliferates and Internet comes up to speed for commercial use, what's next? The concept of broad-band communications always comes back to visual communications, whether you're talking about visualizing data for the computer environment, face-to-face communications or entertainment. For example, ATMs will help open the door to multimedia, because they hold the potential to integrate voice and multimedia video communication.

Companies are introducing plug-in cards for personal computers that mate with the Integrated Services Digital Network and give you a videophone on your computer screen. And one company is already hoping to build on an existing personal-computer infrastructure by mounting a camera on top of its computer.

Business videoconferencing is yet another application of video telephony that's growing strongly. Video-conferencing may not cut down on travel, but it does increase communication. This is another important aspect of multimedia -- it allows people to do cooperative work at a distance, and that's very important for all of us in terms of efficiency. In these systems, the general idea is that you share a video space, much as you would a blackboard. Both people can see the same thing, which you can't do with regular videoconferences. In this respect, it's closer to a telephone conference. In computer-mediated work programs, some of which are just coming to market, you can put things on a computer screen so that everyone can see them, enabling you to discuss documents. You can look at spreadsheets and do design cooperatively, albeit remotely.

These technologies sound promising, especially for companies with employees in distant locations that must work together. But when you look at the way people actually work in their daily lives, it's clear that they're not easily going to fall into working this way. Multimedia technology will have to adapt to their needs.

CONSTRUCTION AHEAD

That brings us to the question of building an infrastructure that will serve users as efficiently as possible. As we increasingly turn to the networks for our information needs, how do we develop an intelligent infrastructure that overlays the mechanics of simple transmission and distribution? Today, engineers and planners are no longer concerned about transmitting data. Instead, they're beginning to focus on building the virtual network that will control the flow of traffic on the highways.

The new problems in network design are constructing intelligent policies and attacking congestion problems. It's the same as designing a real highway. It's not efficient to build freeways that are always free of traffic jams because it costs too much money. You're always confronting the problems of how to let people onto the highway and what to do when they want to exit or too many want to get on somewhere else.

Here's what a typical access problem might look like. Suppose you're sitting in one of these switching nodes and the information packets are flowing in. You've got congestion problems. First of all, each packet of information looks the same. It could be a chunk of speech, part of a picture in a television signal, a computer data file or some multimedia communication. As these anonymous bits start piling up, one packet raises its hand and announces, "I must get through. I'm part of a television picture, and it's going to break up if I don't get through right now. This guy next to me is e-mail. He doesn't have to get through right now, so he should have lower priority."

Scenarios like this mean we have to think hard about access to the highway and which information has the right of way. Plus, digitizing media generates a host of more intangible issues that are ethical and legal in nature, such as copyrights, intellectual property ownership and privacy, and authentication.

Costs are another issue we have to work through, but better technology doesn't mean communications will be free, as some have envisioned. That isn't the way the world works. Even if technology makes the calls themselves very cheap, don't expect communications service to be cheap. That's because the cost doesn't stem from the technology but from the people, the administration, the maintenance and the operations support. Billing calls costs more than all the technology in the network.

And to complicate matters, these anonymous packets of information going through the networks will carry very different price tags, although they'll all look the same. That's because sending video, for example, takes up 1,000 times more information capacity than voice. However, fiberizing the country should make communication costs independent of distance and of bandwidth. That's one of the great possibilities we have before us -- the potential to open up communications by leveraging the information highway's broad bands and user capabilities.

As you can see, the information marketplace is developing and growing tremendously, and our challenge in the political, technological and business arenas is to develop and adopt effective models for the communications infrastructure of the future. It's difficult to predict exactly where all these developments will take us, but it's clear that the information revolution will be user-driven, and that's a departure from the old patterns. Until fairly recently, we were accustomed to developing our communications network through centralized planning, monopolies and regulation. But this time, the marketplace is in the driver's seat.

Mr. Lucky is vice president, applied research, at BellCore in Red Bank, N.J.