*SMARTMONEY.COM: Why You Still Can't Take It With You
August 23, 1999
Dow Jones Newswires
By TIERNAN RAY
Smartmoney.com
NEW YORK -- It's a simple question: Why can't you take your cell phone
with you? It's probably a question a lot of Iridium shareholders and
bondholders are pondering afresh in the wake of that company's
bankruptcy filing on Aug. 13.
The idea of being able to hop from New York to Paris to Tokyo with a
cell phone glued to one ear is - at least right now - purely a marketing
fantasy. For the American traveler, trying to stay unwired in most
European cities involves a series of hassles, including either purchasing an
exorbitantly expensive cell phone overseas or arranging for a rental. I was
reminded of just how relatively underdeveloped our cell-phone system is
on a recent trip to Reykjavik, Iceland. There, the entire population, but
especially young people, are intensive consumers of cell phones; they
swagger around with them on Saturday nights, gabbing with friends in a
price that's generally far higher than the average cost of a cell phone in the
States.
And why not? Europe is intensely wired. One can purchase a cell phone in
Reykjavik (yes, Iceland is a European country) and use it in Oslo,
Stockholm, Helsinki and throughout the rest of Europe. All thanks to the
fact that European phones run a single protocol, or what's technically
referred to as an "air interface," called GSM for Global System for Mobile
Communications. The American market has steadfastly resisted joining
Europe, so American phones use different protocols that are incompatible.
Hence, you can't take your AT&T Wireless or Sprint PCS phone to
Reykjavik, and most European phones won't work here.
This is the whole reason the $1,800 Iridium phone was to be such a hit.
While the I-phone may be big and clunky, it would at least let you take it
with you wherever you go. The problem for Iridium, of course, is that
there are simpler, cheaper ways to talk in New York and Paris than
carrying around a brick that requires special attachments. You can walk
into Omnipoint (OMPT), plunk down a couple hundred dollars, and get a
phone that has radios for two different kinds of GSM protocol: one that
works on the Omnipoint network in New York, and one that will work in
any country in Europe or Asia. Iridium's real usefulness, then, was basically
limited to far-flung parts of the globe - oil rigs, polar-surveying stations -
where there is absolutely no terrestrial coverage.
But the Omnipoint solution isn't perfect, either. Supporting two kinds of
GSM networks takes two sets of chips, but that means something else
ultimately gets left out of these phones. Specifically, they don't talk any of
the other North American wireless protocols, which means that in the
fractured landscape of U.S. wireless services, you risk running into areas
of low coverage. You might have a phone that works in both Paris and
New York, but that goes silent when you go to New England.
Perplexing, to be sure, and it only gets worse as you survey the coming
landscape of cell-phone technology, what's known as "Third Generation,"
or 3G. I've written of 3G recently in this space.
Basically, 3G will provide high-bandwidth wireless connections of 384
kilobits per second on up to 2 million bits. You'll surf the Web, read email,
etc. Fine, except that this new vision brings with it at least three more
protocols for next-generation cell phones (and probably many more
protocols before the thing is baked). Plus, 3G phones in different regions
may "listen" to different pieces of the electromagnetic spectrum that vary
from country to country, and that are different from today's cellular and
digital cellular frequencies. Bottom line: In the next couple of years, you
may have a futuristic, Web-enabled phone - and have an even harder time
trying to communicate in both New York and Paris.
I had hoped that another technology would succeed where Iridium failed.
I had hoped we might see something like a global version of Talkabout,
those cute little blue walkie-talkies that Motorola (MOT) sells at Radio
Shack. True, they won't go farther than two miles, but the Talkabout is the
best commercial example of an old technology that's suddenly important,
something called the software defined radio, or SDR.
Software radio is pretty neat, and it may yet transform the wireless
industry. Sadly, though, it's not going to help the cell-phone situation
anytime soon. Software radio, and a companion technology, the direct
conversion radio, is a way of making a cell phone change its stripes midlife.
Inside every cell phone are two sets of chips: baseband and RF. RF is
what you'd conventionally call the "radio," the ear. It has an antenna to
receive electromagnetic waves, amplifiers to boost the signal, and
oscillators and filters, tuned circuits that listen for the right electromagnetic
frequency. The baseband is the digital brain; its job is to convert the analog
signals from the RF into ones and zeros, and to perform the algorithms that
make the protocols, such as TDMA, work.
The reason a cell phone can't travel is that the RF circuitry is tuned to a
particular frequency, fixed at manufacture for the life of the handset.
Likewise, the baseband only knows how to process certain protocols. A
cell phone in New York might "listen" at frequencies of 869 megahertz and
use TDMA, a protocol that mixes bits in time. A phone in Paris or
Reykjavik would listen at 925 MHz and would interleave the bits to a
different beat, or clock, according to GSM.
You can, of course, build what are called "multimode" phones by
increasing the number of chips - one set of chips for TDMA, one for
GSM. But adding chips makes the phones expensive, and it increases
battery consumption. And sometimes it proves impossible (with TDMA
and GSM, the frequencies are close but not exact, and so it may be
impossible to support both standards in one phone).
Anticipating the increased complexity of 3G phones, chip companies and
carriers and cell-phone makers are exploring software radio. Wouldn't it
be great if, rather than having the RF and the baseband designed to listen
to only one set of frequencies and protocols, the device could change to
hear different frequencies and talk different protocols? The baseband chip
is made up of custom-built circuitry, called ASICs, and math chips, called
digital signal processors, or DSPs. These circuits are set at manufacture. In
software radio, those circuits are replaced or supplemented by
next-generation chips that use some kind of programmable logic. Unlike
ASICs, these chips can be modified while the cell phone is in use, simply
by downloading some software code over the air. Unlike DSPs, they run
really fast.
Texas Instruments (TXN), the leading vendor of DSPs, thinks the DSP has
more life to it. The company's researchers say they've been working on
software radio since the 1980s, when DSPs were used in cockpit radios to
shield them from hostile jamming signals.
The question for the venerable DSP, and for Texas Instruments, is whether
it has the horsepower. Next-generation cell phones will process amazing
rates of data - on the order of 144 kilobits per second and up, as opposed
to today's equivalent of 9.6 kilobits in most digital phones. A modern DSP
can carry out 200 or 300 million instructions with every tick of its clock,
but some estimate that a 384-kilobit data stream would require six billion
operations per second.
The part of software radio where DSPs may prove more useful is
direct-conversion radio . Texas Instruments and Motorola, another DSP
pioneer, along with Analog Devices (ADI), should see lots of opportunity
here. The direct-conversion radio would basically replace the analog
circuits in the RF portion of the phone with some DSP chips. Instead of
circuits tuned to a particular frequency, a set of powerful analog-to-digital
converters, or ADCs, would capture a broad chunk of the electromagnetic
spectrum, digitize it and then use DSPs to pick the desired frequencies.
ADI, some say, has the talent to provide a lot of those ADC circuits.
Long term, direct-conversion radio could kill off large parts of the
analog-chip business. Motorola decided last December to sell off the part
of its business that makes all the analog filters and oscillators for the RF
part of a cell phone, to a chip company called CTS (CTS) of Elkhart, Ind.
Everyone said CTS was getting a real gem, which may be so, but Moto's
view is that over time, the digital chips it makes will replace the analog
circuits it sold to CTS. As one senior Moto engineer tells me, "Over time,
the software radio will replace [analog chips]. Direct-conversion radio is
sort of a quiet war between the digital-semiconductor people and the
[analog] filter people." It's also a threat to companies that make their living
off of exotic RF chips, including one of my favorite companies, RF Micro
Devices (RFMD).
An early example of direct-conversion radio is Talkabout; it uses software
to scan for available channels in the 400 MHz range. That's easy stuff,
though; building a direct conversion radio at cellular's higher frequencies of
900 MHz or 1,900 MHz is a lot harder, and it hasn't really been done yet.
But at least you can feel important when you're walking around the ski
slope yakking on your Talkabout. Unfortunately, none of this will help the
global traveler. Theoretically, a cell phone using a software radio could
change its stripes to suit any network in the world. But that's not in the
cards. Why? It's just not a priority, is the re sponse from Motorola, TI and
other experts in the field. Rather, the focus of software radio will be things
like helping wireless operators stitch together the patchworks of protocols
in their own networks.
So, until some major breakthrough happens, we may still need incredibly
expensive phones, like Iridium, if we want absolute coverage when on the
beach at Odessa, or relaxing by the fire in Bern. Iridium investors, take
heart. There may yet be life in the $1,800 brick.
For more information and a nalysis of companies and mutual funds, visit
SmartMoney.com at smartmoney.com
Copyright ¸ 1999 Dow Jones & Company, Inc. All Rights Reserved. |
