pecific frequencies to each cellular-phone user--cannot keep up with the demand.
The only way to get a tenfold or 15-fold
increase in
capacity is with digital-cellular networks. Virtually every cellular carrier plans to convert its networks to digital technology within the coming decade. But to which standard? Three digital technologies
currently vie
for your company's dollars. The one you choose will affect the scope, reliability, and costs of your cellular applications for years to come.
All three digital technologies are heirs to the Cellular Digital Packet Data (CDPD) technology that's used today. CDPD technology can transmit data at speeds of up to 10 Kbps, sent in the unused channels of existing analog networks. Despite this, CDPD is a greatly underused interim solution. Its three successors will
add better network quality, clearer voice transmissions, higher capacity, faster throughput, and lower power consumption.
Which cellular technology will succeed CDPD? Looking at the three contenders, time division multiple access (TDMA) is a well- entrenched and mature technology with excellent capacity. Second, code division multiple access (CDMA) offers even larger capacity and superior voice quality, even if some of its promised capabilities are still untested. CDMA is also the dominant standard in the new Personal Communications Services (PCS) frequencies (see the sidebar "PCS Goes National"). Finally, Global System for Mobile Communications (GSM) enjoys outstanding international coverage, especially in Europe.
CDPD: Today's Solution
Standardized in 1993 as a bridge to true digital wireless communications, CDPD uses the existing cellular infrastructure as traditional analog advanced mobile phone service (AMPS) in the 800- to 900-MHz range of the spectrum. Because CDPD modems transmit
during idle times--when voice calls are not being made--their impact on cellular voice traffic is negligible, which has given CDPD broad appeal to service providers. That said, data traffic is at the mercy of voice traffic.
CDPD offers other advantages. For example, it uses TCP/IP, which makes any CDPD modem a true IP node, unlike earlier modems, which required a persistent connection, just like a landline-based phone call. With CDPD's so-called wireless IP, modems can receive data anytime they are powered on and within network range, with minimal setup time and connect charges. CDPD's greatest appeal lies in coverage, which includes 40 of the 50 largest Metropolitan Statistical Areas (MSAs) in the U.S. Each MSA represents a region by the population it represents, not its geographical size, so coverage of only top MSAs can still leave gaps in sparsely populated areas. CDPD also covers international sites: four in Canada, four in Ecuador, one in Indonesia, and two in Mexico. CDPD operates not only over
AMPS, but also over other technologies, which lets users share frequencies, including CDMA and TDMA.
There's a broad range of development tools for CDPD, such as Unwired Planet's UP.Link Internet access software or its Handheld Device Markup Language (HDML) 2.0 specification, released in May, for developing applications. Cellular phone/modems cost under $500. For businesses that need digital wireless communications now, CDPD is available. It works, and declining costs make it a feasible technology.
As a packet-based service effectively limited to 10 Kbps, CDPD is best suited to transferring small files, rather than browsing the Web, using workgroup applications, or remotely accessing a workstation. Because of this limitation, in 1996 the CDPD Forum industry association developed a circuit-switched specification for CDPD, called CS-CDPD. Now available, CS-CDPD modems function as wired modems--they can set up dedicated calls. Recent cellular-network enhancements are pushing CS-CDPD throughput to a l
evel comparable to landline connections, though actual performance varies according to carrier, geographic location, and time of day.
Currently, Ameritech, GTE Mobilnet, Bell Atlantic Nynex Mobile, and AT&T Wireless Services have CDPD networks that let customers link in 60 U.S. markets. There are a variety of CDPD modems, from PC Card devices to analog phones with digital modems to hand-held computers. Transmission costs have come down to 8 to 20 cents per kilobyte, on average; most carriers vary prices according to volume.
TDMA: Tried and True
The best advantages of digital-cellular technology are manifested in TDMA: at least three--and in one case 15--times the capacity of analog networks, voice encryption, caller ID, and text messaging. TDMA systems are being designed to operate with 14.4-Kbps data transmission, simultaneous with voice.
To these features, the latest version of the TDMA standard, IS-136, adds extended battery life (by allowing user devices to "sleep" between
transmissions), over-the-air activation, and, according to its proponents, better voice reproduction than its competitors.
TDMA offers Short Message Service (SMS). With SMS, you can switch off the phone, yet messages still arrive, waiting on the network, presented to the subscriber on the phone's display the next time the phone is switched on. TDMA even allows advanced voice-mail systems to send users SMS pages. The information could be the weather, traffic information, taxi information, or other useful local information.
In addition, TDMA's alphanumeric paging is two-way, with the help of a keyboard or stylus on a personal communicator or the alphanumeric keypad on a wireless phone such as the forthcoming Nokia 9000. Subscribers can send messages of up to 160 characters. TDMA offers enormous advances over analog cellular communications to both customers and carriers. TDMA supports seamless roaming between the AMPS and PCS spectra. It will work across carriers, and users might be notified about a
change in billing with a message on their phone's status display.
The IS-136 standard also uses a digital data path called a
digital control channel
, which allows, among other capabilities, the phone to operate in sleep mode. Because TDMA phones are digitally synchronized with the wireless network, phones sleep for all but the 6 milliseconds of every second cycle and awaken when the user receives a call. This extends battery-charge life by as much as 100 percent over analog cellular phones.
For carriers, the IS-136 standard provides for a wide range of service offerings--even for a single subscriber. The phone alerts the wireless network of the subscriber's location. When subscribers are within their office area, the wireless network may give them PBX functions (e.g., extension dialing and conference calls). When subscribers leave these areas--or after a certain time of day--their phones may lose these advanced features and function as any analog cellular phone. Billing could automatically
switch over as well. The phones can even display location identification on their screens, detailing the current billing rate.
Hughes Network Systems enhanced TDMA with its own version, which is called Extended TDMA (E-TDMA). This version more than quadrupled the capacity of the original specification. E-TDMA networks may be able to carry 15 times more mobile traffic than today's analog carriers.
Hughes accomplishes this through digital speech interpolation, in which the transmitter is turned off during pauses and while the user is listening--as much as 60 percent of the total channel time. This idle capacity can be reassigned to other users. Incidentally, E-TDMA networks are backward-compatible with the existing TDMA infrastructure, so the upgrading shouldn't interfere with current service.
The technology behind TDMA has been around since 1967 and is well tested and well understood. It enjoys the staunch backing of LM Ericsson, which is the manufacturer of 40 percent of all cellular equipm
ent sold worldwide and 30 percent of all cellular equipment that's used in the U.S.
TDMA technology is currently embroiled in a small controversy that could become a regulatory petard. Apparently, time-division-multiplexing devices can cause interference with a wide range of electronic devices, from hearing aids to camcorders, laptop PCs, automobile air bags, and pacemakers. GSM devices seem to be even worse offenders. But this is not so with CDMA phones and modems. At press time, a spokeswoman for the FCC had no comment about this, though she confirmed that the matter was under investigation.
A lesser problem has been with audio quality. Early tests over TDMA networks proved to be disappointing. Even analog phones rendered better voice reproduction. TDMA supporters say that these problems have been largely eliminated, thanks to the IS-136 standard. CDMA's proponents, Qualcomm in particular, insist that their technology is far superior in terms of audio reproduction. However, until there's nationa
l coverage with millions of subscribers using CDMA networks, the claims are unsubstantiated.
The more important hurdle is whether TDMA networks will be able to provide adequate coverage fast enough to overcome CDMA's onslaught. Currently, AT&T Wireless is the leading carrier. By the end of the year, the company expects to provide a coverage area that includes almost 80 percent of the U.S.'s populated areas, with 41 of the 49 largest cities.
CDMA: The Technology to Beat
It would seem that TDMA would have the market wrapped up were it not for CDMA, which is purportedly twice as efficient and promises to deliver even better voice quality. CDMA has been gathering slow but important momentum in the past two years, as evinced in part by last year's PCS auction.
Of the 2958 PCS licenses that were awarded in the U.S. last year, 51 percent chose CDMA (with 28 percent going to GSM and 20 percent going to TDMA, according to Edge, the research publication of AT&T). This larger number o
f licensees means that both consumers and enterprises will be able to choose from a greater selection of carriers. This, in turn, should provide the kind of price competition that will make the technology imminently appealing.
Even so, CDMA is a new technology. Many of its capabilities are untested. Others were exaggerated in its early going, which has tended to make carriers suspicious of it.
Dataquest predicts that the number of CDMA users will grow to 60 million by 2000. Some $6 billion has been poured into the technology's infrastructure over the past two years. At least six vendors have committed to manufacturing handsets and modems for the standard. CDMA will putatively deliver all the benefits of TDMA--higher capacity, extended battery-charge life, SMS, over-the-air activation, 14.4-Kbps data transmission, and improved audio reproduction--and then some.
Like TDMA, CDMA features seamless roaming not only between carriers, but also between PCS and traditional cellular systems. However
, supporters say that its network capacity could be 10 to 20 times higher than an analog cellular network's. Supporters claim that it delivers a clearer signal than TDMA, that it entails lower infrastructure costs, and that users would experience fewer dropped calls.
Has CDMA lived up to its billing? Critics, including Bill Frezza, president of Wireless Computing Associates, a Yardsley, Pennsylvania-based consulting firm, have warned that the theoretical potential of the technology would never bear out in the field. Interference would limit its purported capacity, networks under heavy loads could become unstable, and infrastructure costs would be far higher than initially projected.
In practical terms, the technology has stumbled. Construction snafus and lawsuits between several key players have delayed deployment of CDMA networks well beyond early forecasts. Early field tests of CDMA failed to sustain its claims of superior audio quality and network capacity. Meanwhile, field tests of TDMA's netw
ork capacity now approach a ratio of 7 to 1 over analog technology.
Blame the CDMA's youth and inexperience for these blunders and oversights. Introduced with exuberant claims by Qualcomm in 1989, the technology caught the fancy of engineers and carriers because it allowed multiple users to share the same channel.
The costs that both enterprises and consumers will pay for digital-cellular service obviously depend on coverage and competition among the carriers. CDMA now has the advantage on the latter, thanks to the broad appeal of its higher capacity and better voice quality. Several disappointing cases may slow its implementation.
AirTouch Communications, one of the first cellular carriers to commit to CDMA, began plans for upgrading its congested analog system in Los Angeles in 1989. Construction was delayed until 1994, with the planned commercial launch scheduled for 1995. Implementation was plagued by interference from analog service, power- balancing problems, and poor voice quality. By
late 1996, AirTouch had 250 cell sites on-line.
Worse, CDMA network capacity ultimately may reach only six to seven times that of an analog network--which is roughly the equivalent of today's TDMA networks. CDMA's vocoder, the technology that turns analog speech into digital signals, must be set at 8 Kbps to achieve this capacity. That rate is decidedly inferior to the 13-Kbps standard CDMA has adopted.
AirTouch's experiences haven't been enough to frighten carriers away from CDMA, though a string of internal skirmishes have slowed the technology's march. Nokia, Lucent Technologies, Sony (outside the U.S.), Oki America, Motorola (due out in the U.S. by the time you read this), and Samsung all currently offer or plan to introduce CDMA handsets and modems within the year.
However, Qualcomm and Oki are embroiled in a lawsuit (with Ericsson, which does not make CDMA handsets) over patent-infringement claims, the outcome of which could ultimately bear on devices made by other manufacturers. In a
highly publicized reversal, Motorola has dropped its plans to supply base stations for Sprint because of financial terms. Qualcomm and Motorola also filed patent-infringement suits against each other.
Even so, all but one major manufacturer have made commitments to making CDMA devices (many also make user devices for TDMA). A joint venture between Qualcomm and Sony alone could deliver as many as 3 million handsets a year. The company even boosted name recognition recently by paying $18 million to rename the home of the San Diego Padres and San Diego Chargers Qualcomm Stadium.
CDMA also enjoys an enormous advantage in the Block C PCS spectrum (see the figure
"PCS Block Party"
), thanks largely to NextWave Telecom, which was the largest bidder in last year's auction. NextWave plans to build a nationwide "carrier's carrier" network and maintains that it has already sold 10 billion minutes of use to MCI Communications. The company is committed to CDMA.
Lucent Technologies,
which claims to be the largest CDMA vendor, has begun an aggressive vendor-financing package, whereby it effectively underwrites the cost of building the network.
While impressive, this kind of industry backing doesn't ensure that CDMA will become the dominant technology. Coverage will. According to the CDMA Development Group, CDMA carriers cover 100 percent of the U. S. Japan recently committed to the standard. Hong Kong and Seoul, South Korea, also provide coverage. Outside the U.S., TDMA still enjoys better coverage. And GSM is unbeatable. But within the continental U.S., CDMA is clearly the market leader for digital-cellular communications.
GSM: Is Europe's Standard Coming to America?
As pricing goes, enterprises and consumers alike stand everything to gain from the PCS spectrum, which will put enormous competitive pressure on cellular carriers. GSM, which uses TDMA technology in the 900-MHz spectrum, enjoys a larger market share worldwide than either TDMA or CDMA--some 70 countries a
nd still growing. With the auction of the PCS F-block last year, the standard, now encompassing PCS-1900, is now poised to offer 80 percent coverage of the U.S. as well.
What's more, it can provide that coverage far faster than its leading competitor, CDMA. GSM technology already owns a large share of the world's wireless user devices, second only to pagers. Carriers are attracted to the technology in part because vocoder standards and voice-compression algorithms have been perfected since the mid-1980s. Transferring GSM to the North American 1.9-GHz spectrum promises to be both quick and relatively inexpensive. GSM carriers who paid exorbitant sums for their PCS bandwidth can expect a shorter time to a positive cash flow than CDMA carriers.
Ericsson, Mitsubishi, and Nokia already provide GSM handsets. Various manufacturers supply PC Card modems. The technology can already deliver to users the same benefits you'd get from CDMA and TDMA technologies, including better voice quality, higher network c
apacity, SMS, data and voice encryption, international roaming, and billing flexibility. It can also simultaneously process voice and 9600-Kbps data. It is not surprising that Pacific Bell, BellSouth, American Personal Communications, Go Communications, and Microcell 1-2-1 (Canada) have chosen the technology over CDMA.
Pacific Bell, which has inaugurated service in California and Nevada, has become a showcase for the technology's potential. When subscribers press the send button on their Pacific Bell handsets, they're automatically connected to a customer-care representative. Subscribers can program the network to have specific calls follow them. They can store both voice mail and pages in a universal wireless mailbox. SMS service includes multiple languages. Pacific Bell and Ericsson (which manufactures the handset) have agreed to develop an interim solution that would block the interference that people with hearing aids currently encounter.
Eventually, a triple-mode, triple-frequency, triple-cod
ec handset using GSM technology could bring about a world phone, which would be compatible with GSM networks at 900 MHz and the DCS-1800 networks at 1.8 GHz used outside the U.S., as well as the PCS-1900 network used in North America. In effect, it would be a world communicator, which could be carried and used anywhere in the world.
Meanwhile, GSM continues to enjoy double-digit growth everywhere outside the U.S. and Japan. Some estimates give it 50 percent of the world mobile-phone market by the year 2000. Thus, it should take hold in the American PCS market long before its biggest competitor, CDMA, is able to get a footing, and even before TDMA networks are up and running. Over the long haul, it is the single best solution for worldwide roaming.
On the other hand, GSM can't match CDMA's purported network capacity. Many analysts are skeptical about GSM North America's claim of 98 percent coverage in the U.S. This has prevented more than a few carriers from signing on to the technology. CDMA still
has the edge in North America. For most enterprises--even those with manifold international offices--universal coverage in the 48 contiguous states is of paramount importance in building wireless applications. For most U.S. organizations, the rest of the world can wait.
Your Best Choice
Can TDMA, CDMA, and GSM coexist? For the foreseeable future, there's no question they will. Sooner or later, however, one will predominate, ushering CDPD and analog cellular communications out of the picture altogether. But betting on one technology over the other today is a precarious gamble. This is especially true for enterprises making broad-based and long-term investments in developing wireless applications. For all practical purposes, the troika of digital standards will continue to battle each other well into the next century.
Besides, venturing into a true digital world may be premature. CDPD has much to offer companies that need to create wireless applications today. It doesn't require a large ca
sh, time, or technology investment in standards that could be swept away in the next five years.
The choice between TDMA and CDMA is more difficult. TDMA enjoys broad coverage throughout the U.S. and, to a lesser extent, in Europe and Asia. It is a well-known technology in terms of its capabilities and limitations and, more important, in terms of costs. Some incremental improvements--especially the advent of Hughes' E-TDMA--make it highly competitive with CDMA, even if both carriers and user-device makers are marching en masse toward the latter. (Don't be fooled: Both contingents are ready to jump ship if the other standard prevails.)
CDMA has real technological advantages--especially in terms of future expansion of networks. This explains why so many carriers are infatuated with the technology. The more, the merrier, since the resulting competition will push subscriber costs downward. However, the technology has gotten off to a slow start. Its much-trumpeted potential hasn't yet borne out in ear
ly implementation. Industry infighting continues to plague it. The digital wireless war is CDMA's to lose. It's too bad the technology has thus far been its own worst enemy.
GSM is the wild card in this contest. Widely used in Europe and other parts of the world (including many places where it is cheaper than copper wiring to set up and use), it should be your first choice for wireless applications on an international scale--at least for today. Even in the U.S., it has an aggressive presence in the PCS spectrum. Also, it's based on the same time division technology as TDMA, which means that its capabilities and limitations are well known. Coverage in North America is good, if not complete. But can GSM entice enough carriers to its standard so that pricing would be pushed to competitive levels with CDMA? Can it build adequate capacity so that your enterprise can depend on it?
All these unknowns reinforce the fact that digital wireless technology, though no longer inchoate, still has a long way to g
o. There's far more certainty about the kinds of services and applications that you can expect than there was five years ago. There are also a world of ways you can harness the technology today. But reaching the point where you have a single reliable standard served by multiple carriers is still several years away. Getting there is going to be very slow indeed.
Where to Find
AirTouch Communications
San Francisco, CA
Phone: 415-658-2000
Internet:
http://www.airtouch.com
Air War
