C2, I doubt that QUALCOMM has much of a position in OFDM. Several years ago [about 1997 or maybe 1998], I hunted down some OFDM research at Auckland University [I had never heard of it]. Clark Hare and others who know stuff climbed around the idea and we had Andrew Viterbi asked about it after an AGM [I stood by while he answered]. He said that QUALCOMM has a few patents around it or some off-hand comment. He seemd to me to be avoiding the question and minimizing the matter.
BUT. Not long later, what a surprise, he left QUALCOMM and joined Flarion. So, he obviously thought OFDM was worth backing in a big way. flarion.com
Meanwhile, OFDM was trundling along in the WiFi world.
There is no indication at all that QUALCOMM has any kind of position in OFDM patents requiring merchants to pay tribute to Mighty Q. Therefore I conclude that the patents are perhaps useful in some respects, but no big deal.
Flarion has got their technology up to commercial stage; about where QUALCOMM was in 1995 when various pre-commerical trials were being conducted.
802.11g and other OFDM commercial technologies are available.
Flarion's technology seems to be something like 3 times as good as QUALCOMM's CDMA for spectral efficiency for data, but for voice, it is perhaps no better, which I suspect is where Irwin Jacobs and Andrew Viterbi parted company.
Andrew Viterbi seems to have been the data point-man and Irwin Jacobs the voice enthusiast. Those two conflict and QUALCOMM resolved the matter with 1xEV-DO [aka HDR] with channels being allocated to data or voice depending on instantaneous demand.
Since Telecom New Zealand is selling 1xEV-DO data at US35c a megabyte [if you agree to buy a pile of them every month whether you use them or not, meaning the true price will be about US70c a megabyte] and the production cost is something like US2c a megabyte [according to a years-old white paper from QUALCOMM and I suppose those costs have come down a lot since], it won't be a big deal if Flarion can squeeze 3 times as much through the spectrum.
That's the same reason GSM retained its advantage over CDMA despite the huge spectral advantage of CDMA [something like 10x as efficient]. The retail price of minutes was so high that the savings in spectrum was trivial compared with all the other costs in delivering a minute to the subscriber, not to mention the quality and price of cellphones - GSM was much better.
So, since Flarion's OFDM's advantages are small compared with CDMA, which has got a bigger head-start on Flarion than GSM had on CDMA, I can't see Flarion or OFDM in wide area networks getting a foothold any time soon.
But in the WiFi range, up to 100 metres, I think OFDM will have a LOT of fun. WiMax too will have its day, enabling WiFi hot zones to be linked together easily.
Irwin Jacobs is very much a customer-focussed person. Many times I've noticed him take a question back to the driving force, which is the experience of subscribers and what they will want. They are the ones who pay the money which makes the whole industry go. It's a shame the Globalstar people didn't know that.
So, OFDM isn't a big worry. Yet. But, maybe there really is a 10x advantage over CDMA. Which then requires an answer to the question of just what is the value of spectrum?
The value of spectrum is a balance between the cost of hardware and demand for spectrum. If hardware was free, service providers could build base stations every few metres and that would mean frequency could be re-used every 100 metres, instead of every 5 kilometres. That would be equivalent to a vast increase in spectrum availability.
Hardware isn't free, so there's an economic balance between base-station spacing, spectrum, demand for data and the price charged to subscribers.
As the cost of hardware declines, that reduces the advantage of OFDM over CDMA since more hardware can be installed to solve the spectrum shortage problem.
I suspect that human brains are too small to use all the spectrum that's available even with current hardware costs. As costs continue to decline, I suspect there will be surplus spectrum.
But, I'm watching this space. OFDM might spring some nasty surprises. WiFi could gain a LOT of ground.
Mqurice
PS: Good old Google has a link [better than SI's search function] Message 14121003 |
