EricL, I missed the transition from cdma2000 to CDMA2000. Thanks for the information. I used to wonder whether cdma2000 should be Cdma2000 at the beginning of a sentence.
There's no spectrum constraint in most places. In Auckland, where I offered lots of money to the government for spectrum for all of NZ, the value is really just to stop competitors getting into the act and to maintain an oligopoly between Telecom and Vodafone.
We [son and I] tried to figure out the spectrum value and it really is near zero. That was reflected in 2GHz bids. The government thought they were going to get heaps of loot following the European auctions, but by the time NZ got into gear, 3G hopes had dwindled dramatically. Now, the value of spectrum has dropped further.
As the cost per kilogram of electronics continues to drop and the brainpower of it continues to rise, the value of spectrum will fall further.
Civil engineering costs can be avoided by infilling with hordes of microcells. Just glue electronics on any vacant power pole or even sprinkle them from top-dressing aircraft, with only the lucky surviving, a bit like DNA is sprinkled into the biosphere and the lucky and smart survive. If it's cheap enough, sprinkling is the cheapest way of doing it.
We're a long way from that, but dense infilling is not too expensive now. Vodafone is doing it with GSM on each street corner and on each floor of some buildings. GSM has done a great job of holding the line against CDMA with continuing technological advances and cost reductions.
But they can't beat physics or economics. In densely populated environments, spectrum is still valuable and sprinkling picocells isn't economic. GSM is like Al Qaeda and SARS, it just keeps coming!! Though AlQ hasn't had a high profile lately.
I don't see why anyone would use 2GHz when they have plenty of 800MHz to backfill with more efficient CDMA2000 and multimode handsets [GSM/CDMA]; yes, when they are finally produced.
Vodafone in NZ has heaps of spectrum around 900MHz. They just bought more [890 - 899, which I wanted]. I suppose they are planning to fill it GPRS. But GPRS is not doing well. I suspect they'd do better to plug in some GSM1x at each of their base stations and sell multimode handsets. They do have 2GHz spectrum in which they were going to build W-CDMA. But they won't.
We worked out our costs in 2GHz and 890MHz and you just don't do it in 2GHz if you can do it in 890MHz paired with 850MHz. The capital cost, due to extra base stations, for 2GHz would kill it!
There is never going to be W-CDMA in New Zealand. Nor in similarly less densely populated cities. Auckland has 1 million people and is somewhat like Los Angeles in population density. A high-rise core and then single storey as far as the eye can see. But LA goes further than the eye can see.
I doubt that places like London, Los Angeles and so on can justify W-CDMA in 2GHz instead of CDMA2000 in 800MHz or 900MHz, or better still, 450 MHz.
Spectrum should be used for the most economic purpose and cyberphones are the most valuable purpose going! A few low-rent people wasting a bunch of space in 450MHz while umpty $$billions is wasted to build out 2GHz is insane.
In Shinjuku, Oxford Street, Ginza, and Manhattan, where cyberphone toting humans will be hexagonally close-packed, it might make sense to build W-CDMA because numbers will dictate close-packed base stations and 2GHz might even be more desirable than leaky 800MHz or 450MHz which will propagate for miles. That's engineering detail I don't fully understand, but maybe that would make sense.
Despite asking and puzzling for years, I am still unable to give any good reasons why W-CDMA is preferable to CDMA2000, especially now that CDMA2000 has capital letters too. And especially when W-CDMA in 2GHz is considered instead of 800MHz or 450MHz.
Mqurice |
