The Need For Speed

Waiting for the wireless data market to take off is like waiting for the Chicago Cubs to win the National League pennant. At the beginning of each season, loyal fans are full of hope. But as the season draws to an end, it becomes clear this was not meant to be the year.

Almost everyone has an opinion about what it will take for wireless data to be a winner. Lower prices, more end-to-end solutions and ease-of-use frequently show up as areas needing improvement. Another recurring theme is "lack of customer awareness."

Now, there is reason to believe wireless data really can be a winner. What it lacked for many years is the same thing the Chicago Cubs lacked: performance. Years of failure -- trying to convince users to dumb-down their data applications to work over slow-footed wireless networks -- have forced the industry to take a new tack. A bevy of high-speed wireless data solutions are about to take the field.

Picking Up Speed Why is fast throughput so important? Thanks to the World Wide Web, we now expect networks to deliver multimedia content. So to be a contender in the remote-access race, you must perform at least as well as today's fastest dial-up modems. (Dial-up modems, and the RJ-11 jacks they plug into, have become ubiquitous.) Not surprisingly, customers stay away in droves from wireless networks that run noticeably slower.

The 1999 wireless data season looks particularly exciting because there are several fine prospects. Competition in the wireless industry is spurring vendors to develop third-generation (3G) solutions capable of delivering up to 2Mb/s to individual users. Fixed wireless operators are being armed with 200MHz to 1,000MHz of bandwidth in LMDS, wireless cable and millimeter wave bands. Wireless LANs running 1Mb/s to 10Mb/s finally are poised to take off. And early in the next millennium, exotic new satellite networks promise to upgrade the entire planet's telecom infrastructure.

The dial-up modem has come to its evolutionary dead end. Yes, you can squeeze 56,000b/s out of a standard, 3,000Hz-wide telephone channel, but only under optimal conditions. The most promising wireline alternatives are ISDN, xDSL (a family of digital subscriber loop technologies) and cable data. But each has its coverage limitations. ISDN and xDSL work over regular telephone lines -- as long as they aren't too long and don't have loading coils or bridge taps. Most cable TV networks can transmit data to users, but only a fraction accepts upstream transmissions, and even fewer provide service outside residential areas.

Cellular/PCS Market The potential market for digital cellular/PCS-based, high-speed wireless data is huge. The current obsession began when Japan's Nippon Telegraph and Telephone (NTT) announced plans to deploy a 3G wireless telephone network. The basic capabilities of 3G wireless networks have been defined by the ITU's IMT-2000 initiative. Namely, 3G networks must support 128kb/s for vehicular mobile users, 384kb/s for pedestrian mobile users and 2Mb/s for fixed users.

What makes 3G wireless so attractive is it focuses on the next step up, speed-wise, from dial-up modems, yet it doesn't attempt data rates so high they quickly sap network capacity and coverage. As the Internet continues to evolve, demand for faster access is sure to grow. Plus, wireless offers the most convenient method for exchanging e-mail and surfing the web from airports, hotels and convention centers.

Like local telephone and cable TV operators, wireless operators entering the data market enjoy significant economies of scope. For example, they can use the same infrastructure to serve both the voice and data markets. But unlike local telephone and cable TV operators, wireless operators don't have to run new cables or upgrade old ones. Thus, 3G wireless operators can provision service quickly and may even help local telcos and cable TV networks extend the reach of their high-speed data offerings.

But the digital cellular/PCS industry is mired in politics. Every major player now agrees that 3G solutions will be based on some form of CDMA. Ericsson has proposed and is developing a wideband CDMA (W-CDMA) that is not backwards-compatible with cdmaOne as used in the United States, Canada, Japan and many other countries. However, Ericsson claims its W-CDMA has technical advantages (disputed by cdmaOne proponents that call their 3G solution cdma2000) and is designed to work with the evolving GSM core network. Key differences between W-CDMA and cdma2000 are their chip rates and use of asynchronous or synchronous base stations.

Globally, Ericsson has won more support for its W-CDMA (even Japan and Korea have endorsed it), but Qualcomm holds key intellectual property. Still, it remains to be seen whether W-CDMA, cdma2000 or both can deliver reliable high-speed data with abundant capacity at prices users are willing to pay. So far, manufacturers have done little to identify, quantify and characterize the market for 3G wireless.

Millimeter Wave Services But 3G digital cellular/PCS is not the only high-speed wireless data game in town. LMDS and other millimeter wave services have been allocated huge bandwidths. LMDS A-block auction winners, for example, are licensed to use an unprecedented 1,000MHz of spectrum. (That is more bandwidth than was contained in the entire commercially viable radio spectrum just a couple of decades ago.) LMDS equipment manufacturers plan to support services ranging from 16kb/s to 620Mb/s.

Although LMDS (28GHz band) and other millimeter wave services (24GHz and 38GHz bands) eventually could provide high-speed access for hordes of Internet users, most observers expect operators initially will target small- to medium-size businesses. One limitation of operating at such high frequencies is the need for line-of-sight access. That is not a problem in most business districts, but it is a challenge in residential areas, where tree foliage obscures a substantial percentage of homes.

The most exciting development in the LMDS/millimeter wave bands is the use of point-to-multipoint networks to deliver an array of telecommunications services. Although traditional telecom networks use dedicated, point-to-point links, the new broadband wireless networks can employ more cost-effective, shared links. Nationwide operators such as Teligent, Winstar and Advanced Radio Telecom believe they possess technologic and regulatory advantages over the incumbent local exchange carriers (ILECs) -- at least in serving business customers. Look for these new wireless operators to steal many of the ILECs' best customers.

Other wireless operators are concentrating on high-speed Internet access for small businesses and telecommut-ers. Wireless cable operators, which use the 2GHz multichannel multipoint distribution service, multipoint distribution service and instructional television fixed service bands, but have enjoyed little success in the cable TV market, now are flocking to the Internet access market. Some offer their wireless bandwidth exclusively for downstream transmissions, requiring customers to use dial-up modems for the upstream path. Others offer 2-way wireless transmission using the nearby wireless communications service for the upstream path, but at hefty prices -- from $150 per month to well over $1,000 per month, depending upon the data rate subscribed.

Wireless LAN Meanwhile, new companies applying wireless LAN technology to the high-speed Internet access market are springing up all over the United States and Canada. The basic idea is sound, but the technology and regulations probably need fine-tuning. At best, these firms can provide shared access to wireless links running 1Mb/s to 2Mb/s. Thus, users can expect fairly high burst speeds, but average throughput may be much lower, depending on how heavily loaded with users the networks are permitted to get. (It is even conceivable that subscribers on poorly managed systems will see average throughputs less than what they could expect using dial-up modems.) Wireless LAN technology also is being used successfully to provide high-speed point-to-point links in developing countries where wireline infrastructure is lacking.

Satellite Within five years or so, broadband satellite networks will bring modern telecom services to almost every point on the earth's surface. One of the most ambitious projects is Teledesic, founded by Craig McCaw and Bill Gates and into which Motorola's Celestri program has been folded. Teledesic will use 288 satellites in low earth orbit (LEO) to deliver services ranging from 16kb/s to 2Mb/s (plus asymmetric services up to 64Mb/s). Teledesic's use of LEOs will help keep subscriber terminal prices low and ensure terrestrial-like propagation delays. Other broadband networks will use satellites in much higher geosynchronous orbits, which will require subscriber terminals with more powerful (and subsequently more expensive) transmitters.

The sheer number and variety of solutions being offered, or about to be offered, for high-speed Internet access suggest wireless data is about to become a serious contender. Unlike wireline operators, which must run dedicated cables to each business and residential customer, wireless carriers can offer true bandwidth-on-demand service. Plus, only wireless operators can offer high-speed services that are portable. (Don't even think about doing that with your cable modem.)

Most people thought wireless data would succeed because it was the only solution for mobile users. But it is beginning to look as if wireless data will succeed because, given all of its new bandwidth, wireless networks now can provide more cost-effective, shared, high-speed access. Mobility, when supported, is simply icing on the cake.

So next year looks pretty darn good for wireless data. But what about the Chicago Cubs? Well, next year looks pretty good for them, too.