Wireless Pioneer on Europe 3G and CDMA
Qualcomm Founder Reflects On The Past, Looks To The Cdma Future. By Marisa Torrieri
(Thanks to Phillips Telecon)
You hear some version of the following story a lot these days: seven guys in a small shop,
working round the clock to make their technological dreams into a money-making reality. But to
think that 15 years ago, that was wireless giant Qualcomm's [QCOM] (more) beginning, is
quite another thing.
Andrew Viterbi, a pioneer of CDMA technology, and co-founder of Qualcomm, recalls those
first days when the wireless startup had less than 10 employees and was more of a great idea
than actual reality.
In this exclusive interview with Wireless Data News, the former Qualcomm vice chairman and
world-renowned electronics engineer discusses his views on CDMA, his so-called retirement
and predictions for wireless data:
WDN: We've heard stories of working in the garage. What were the early days at Qualcomm
like?
VITERBI: We had seven people, and it was not in a garage but in a loft, and it was on the
second story of a commercial building where the ground floor was a cleaner and later became a
pizza joint. We were only there for three months while we were waiting for our first permanent
building, which is actually still in the Qualcomm inventory. It was very exciting - the spirit was to
figure out from our past experience what kind of business we could go into and certainly it was
going to be telecommunications. At that time, wireless as a commercial business was hardly a
blip on the horizon. Our first customers were military satellite people primarily, but along the
way, what is today Omnitracks became an opportunity, so we actually started working on that
in the loft during that first three-month period in 1985.
WDN: What are you going to miss more now that you're retiring? Your academic career? Your
position as vice chairman of Qualcomm?
VITERBI: I think perhaps the wrong impression has been given. Yes, I'm retired from my
executive position at Qualcomm. I'm still on the board at Qualcomm; in fact, we just had a
meeting. And furthermore, I'm not sitting on the beach even though the weather is beautiful
today. I've been extremely busy, really, with travel. I'm giving some half-dozen or more lectures
all over the world between my retirement and July at various places - universities, USC,
Princeton, MIT, and I'm still quite active academically and professionally. I'm also sorting out a
lot of people approaching me - these are mostly investment opportunities. I'm on half a dozen
technical advisory boards in a variety of areas ranging from Web-based learning to some
software Internet companies to some wireless. And I will have an office to keep all of these
activities sorted out. So, I'm still a very active 65-year-old.
Certainly I found five years ago I was getting too busy to teach a regular course at a university
which I had been doing for 31 years. The last 19 through [first company] Linkabit, and let's say
the first 10 years of Qualcomm, I was actually a part-time professor, but I did lead a class
regularly for about one-quarter of the academic year. But really the international travel and things
that required my being away made it very awkward to continue, and that's why I retired from
teaching at the university in 1994. So I miss that to some degree, but I'm so busy with other
things, including giving a lot of lectures and being involved in a number of technical advisory
roles, that I can't say that I'm bored.
As far as day-to-day management, that was never my forte. It was not a dramatic change going
from academia in 1973 to our first company.
WDN: Do you envision writing papers and textbooks in addition to your CDMA textbook?
VITERBI: I'm still giving talks, and they range from overall industry surveys to highly technical
and mathematical papers. That's not my primary activity, but I'm not turning that off either. Yes, I
will continue to do some of that; whether I will revise the CDMA book, which I've been asked
to do, I think I'll let things settle a little more before I begin to tackle that. But there's a good
chance I might do that in the next five years. I've also been asked to write a section for a
telecommunications encyclopedia. It's not that I retired in order to do that, but I'm not turning it
away either.
WDN: What motivated you to leave the world of full-time academia for the more managerial
functions you performed at Linkabit and Qualcomm?
VITERBI: First of all, the motivation was the excitement, actually, seeing some of that academic
work become a reality and it became much more so than I ever expected. Of course, it took
about 20 years for it to happen, but it's been an exciting 20 years of evolution, I'm very, very
satisfied with what I've been able to do because of leaving academia then.
There certainly was managerial - for a while I was president of Linkabit, and I've certainly taken
an active managerial role, but I don't consider that to be my primary activity. My primary activity
at Qualcomm was in keeping our technology fresh and recruiting the very best people and
maintaining the contacts in the profession and in academia that would help us recruit the very
best people, which we were very successful at doing. I don't take full credit for it by any means,
but that was a major part of my duties and I certainly enjoyed that.
It wasn't a dramatic change, whether back when I left UCLA in 1973, or as we started
Qualcomm. It was a very technical activity, and better than half my time was on technical
matters.
WDN: When you began developing the CDMA standard, did you envision that wireless would
become such a part of everyday life?
VITERBI: We began the CDMA development in 1989. It was already clear that cellular
telephony was going to play a major role in peoples' everyday lives. But at that time, there was
perhaps the penetration of 10 to 15 percent in the United States and probably less elsewhere,
even in Europe. Since that time, Europe is up to almost 50 percent and the U.S. is getting close
to 40 percent and this is only the beginning. That's just in the perennial voice conversations. The
future of wireless data makes it such that we're going to have virtually 100 percent penetration or
more in the sense that you'll have several devices talking to each other, talking to machines, or
machines talking to machines. No, I had no idea in 1989 - I or anybody else in the company -
that it would reach these proportions.
But I will say that we did recognize in our clients - in our initial backers among the service
providers - that they would need to have a much more efficient system, that they would make
better use of the spectrum they had by at least an order of 10-to-1 in order to grow to support
even what they predicted at the time. And that's why CDMA got started and gained its backing.
WDN: What role did the introduction of your Qualcomm CDMA phone in Korea play in its
ultimate success in the United States?
VITERBI: Well, Korea was the first country to standardize [its] digital cellular and digital PCS
on CDMA, and in fact, exclusively on CDMA. Their analog system, unlike the United States,
was somewhat inadequate, whereas ours by 1996, our system was already quite capable and
pervasive.
In Korea, which used it much less, they were having a lot of problems partly with the terrain, and
partly with many other factors that were not in the present United States. By backing CDMA
from day one, and getting a tremendous acceptance because they backed it fully, that created a
tremendous customer base, which until recently, was still probably more that 50 percent of the
world's phones. Today, I think it's considerably less so;, we're somewhere well above 50 million
heading toward 100 million, but Korea still represents probably 30 percent.
Therefore, the take-up was much slower than the United States, until PCS came along, and that
wasn't probably until '97 and '98. On the other hand today, the growth is extremely large, and
since then, Japan has come on board also.
But there's no question that the Korean situation and experience really motivated and
accelerated the use of CDMA worldwide.
WDN: What do you foresee in terms of competition to CDMA?
VITERBI: As you know, for 3G, the vast majority of carriers, governments and standards
bodies are embracing CDMA. There is a slightly different flavor, which I don't think we need to
go into, but it's all CDMA, it's all spread spectrum, and it's all using direct sequence. It's all
fundamentally the same thing, as was IS-95, or CDMA1, in an evolution. And, obviously, in 10
years, you're going to evolve and improve the standard, which is exactly what has been done.
Qualcomm, of course, with support from a number of other manufacturers, including Lucent
[LU] (more) and I believe Ericsson [ERICY] (more), are now evolving the original CDMA1
IS-95 into a further evolution which is now being called 1X. So evolving, we're getting capacities
that are at least double than the original 1S-95.
More important, I think the trend is toward data, and there we're getting even larger beyond that
doubling. We're getting another factor of three to four by recognizing that we have packet data
and making all the allocations based on optimizing the throughput of packet data, by recognizing
there's a strong influence by the environment which you can take advantage of in wireless data
that you can't take advantage of in wireless voice. So we advocate very strongly the separation
of data and voice into separate carriers; a carrier being a one-and-a-quarter megahertz of
spectrum. Others are pushing other ideas which we don't think are nearly as far along.
So, it's going to be some flavor of CDMA, and I think the demands for high-speed wireless
data, which will be competitive in speed with DSL and cable..., [and]Something you can carry
with you, and you're not tethered to the wall. That's the major trend.
WDN: What about any challenges en route to 3G?
VITERBI: Do I see other technologies other than CDMA that are challenging? Frankly, I'll be
very blunt and say that I don't think EDGE [a TDMA technology that's an evolution of GSM]
has very much of a chance. Packet data over GSM, so called GPRS is happening. It's a very,
very slow service, far slower than IS-95b, which is already in operation in Korea and Japan and
shortly in the U.S. - I'm not sure exactly when, but it'll be this year.
There's 200 million people with GSM, and if carriers can entice people to buy new handsets
based on the GSM standard, they can go with packet data. But it'll be a relatively slow packet
data, and it'll be mildly expensive - far more expensive than the CDMA handsets that already in
IS-95a can handle the same speed as GPRS, and 'b' can handle approximately two-to-four
times the speed of GPRS. So, that, I don't see as being any challenge to CDMA in efficiency
and performance. If the Europeans want to spend a lot of money to develop it, that's another
story.
As you know, the Japanese have completely committed to 3G-CDMA for their ongoing data
requirements. The Europeans, however, do have this EDGE technology, which [is a]
technologically I don't think offers much promise. So, I don't see that as a threat.
If you want to go way out, to the people who are promoting fourth- and fifth-generation
ultra-wide band schemes, which require a gigahertz which they hope would become unlicensed -
well, I say that likely, if at all, it's 25 years off. The regulatory issues are enormous, and I think
the technology's probably a decade away from being realizable. So whether that will be
economically feasible, in addition to technically conceivable, who knows what may come along
in the next decade?
Will we have short-range optical? That I can't say, but for the time being, I think wireless is being
very well-served by CDMA for the next decade, easily. |
