April 17, 2005
From the desk of Jane Galt:About that article
Liberals are championing this Foriegn Affairs piece on broadband as if it just goes to show you that the Bush administration is keeping us from the high-speed broadband that is our natural right. If we just had a good regulatory scheme like Japan's, they say . . .
[All right, we'll take a moment so that those of you who know something about the regulatory environment in Japan can wipe the saliva off your monitor. If you've aspirated any Hi-C, make sure you're all right before you continue reading.]
But it's not some indefinable excellence of regulation that caused Japan to get more high-speed broadband; as the article makes clear, they subsidised the hell out of it. The piece is a little bit shifty . . . he keeps talking about "access" rather than uptake, which is the relevant number; if 90 percent of Japanese have been given, at great government expense, something that 1 percent of Japanese actually use, that's not a great deal. Anyway, color me skeptical. Plus, remember how Europe's cell phone network was, like, eighty zillion times better than ours until it turned out they couldn't afford to upgrade to 3G? Big, honking government sponsored infrastructure projects, in which the government picks a technology winner, generally look better than the market's bumbling trial-and-error approach right up until it turns out that the government has made a whomping big mistake and it's too costly to fix.
Plus, he's urging us to make sure that by 2010 there's high speed internet access to every nook and cranny of America for $20-25 a month, which seems to betray a fundamental, total lack of understanding of the difference between Japan and South Korea, which have population densities of 327 and 432 persons per square mile, respectively, and the United States, which has a population density of 27 persons per square mile. Oh, we cluster, but still, all those square miles are between people who need broadband access, like the ranch I spent my summers on in Wyoming, two hours by car from the nearest town and several miles from the next neighbour. And if you've spent time in American suburbs or exurbs, and then gone to Europe or Asia, you know that we really do live differently. That this fellow doesn't seem to realise this makes me suspect everything else he says.
Posted by Jane Galt at April 17, 2005 3:49 PM | TrackBack | Technorati inbound linksUsing WiMax defeats the population density argument.
US communications policy is run for the benefit of large corpoarations. The FCC is bent over the desk servicing Verizon and Comcast. Phone bills run $50 per month, when they should be $5. This is really government-sponsored organized crime.
Twenty years ago I was spending $200 a month on long distance. Our total phonebill isn't going to get down to $5 as long as US communications policy requires me to subsidize service for areas of low population density.
Average total cost, soft and hard, to put up a new cell site is a quarter million dollars. It would be even higher, but there is an FCC policy more beneficial to customers than providers which requires companies to provide co-location to their competitors.
GSM, the standard chosen by the European governments, did not turn out to be a big whomping mistake. CDMA's much-touted spectral efficiency requires a power-density tradeoff; to serve more customers the sites must be closer together. The fact that I work mostly with GSM may be influencing my opinion on this, but US GSM providers are continuing to expand their networks rather than switching technologies. I just passed up a six month GSM optimization contract in South Carolina because I'm not ready for that long on the road and had a briefer offer, closer to home with a better per diem.
One major factor in the different development path for cellular in Europe was that there were far more people who had never gotten landline phones for whom cellular was their first. With in-building coverage expected, far more infrastructure is required. The business model, at least in Italy, is very different, with phone sales completely seperate from connectivity.
Jane, you know full well that the guy writing that article doesn't want to hear about problems with it.
He's done the hard part by figuring out what's needed. It's up to people like you and me to take care of the niggling little details of making it happen, and we're not supposed to bother him with reasons why he can't have what he wants.
Triticale, your information is wrong. If a CDMA network and a TDMA network have licenses with the same bandwidth allocations and have equal density of cells, then the CDMA network will be able to carry about three times as much traffic as the TDMA network.
Actually, when it comes to GSM 2G specifically the ratio is probably even greater. GSM allocates 25 KHz each direction per traffic channel, and these days most CDMA networks are using EVRC and only use 9.6 KHz even when full rate, which they aren't most of the time.
That's one of the reasons why the GSM committee chucked TDMA for GSM 3G and went with a CDMA air interface. (But there are other reasons.)
GSM providers are expanding GSM 2G instead of switching to GSM 3G because there's no organic upgrade path. A 2G handset can't talk to a 3G base station because TDMA and CDMA are fundamentally incompatible, so if they fully transition to 3G they have to completely replace every handset belonging to their customer base.
And a partial cutover has other problems which are just as bad.
One of the reasons that IS-95 CDMA providers are upgrading to CDMA2K is that the upgrade is much easier. An IS-95 handset can talk to a CDMA2K base station, so the infrastructure cutover can be incremental.
I find it amusing when people talk about the power of the free market when talking about US telecom. Truly, Stalin would be proud of our telecom legislation.
Hint: if Verizon weren't writing our laws, we'd already have quality broadband.
I've worked in this market for 15 years now.
I bow before superior knowledge. SDB's involvement in CDMA is orders of magnitude greater than my involvement with GSM. I have to ask however whether the comparative numbers factor in the use of ASR to split time slots, which will theoretically double traffic capacity once everyone has capable handsets.
Correction. The kludge to share timeslots in a GSM traffic channel is AMR, not ASR.
Ya know, when I first heard about the Philly WiFi project, I thought, "hey, what a neat idea". But then I thought about it; and I wonder - why is the City of Philedalphia going into business to compete with cable, DSL, and high-speed cellular? Why is it necessary that the city run a ISP? At subsided prices to boot...
As a European who's been bombarded with adverts and offers to switch to 3G (and I mean proper 3G, not the GPRS-speed CDMA2000 1X y'all have Stateside) for almost two years, I'm slightly surprised to learn that Europe's telecoms providers can't afford to switch to 3G. Who knew?
if "3G" is the one that gives you video clips of the football highlights (and presumably, pornography) then I've got it. I got it from Orange rather than the helpfully-named "3". They have it in France and Germany too, although apparently takeup is not that great, mainly because it's a bit of a waste of money.
CDMA has had working "3G" service (EVDO) since at least the middle of last year - it's been slow-rolled here because most people don't seem to care about data on their phones. There's a cultural difference right there.
"Hint: if Verizon weren't writing our laws, we'd already have quality broadband."
We _do_ already have quality broadband, in every place where supplying it is economically efficient. And getting it in places where it _isn't_ economically efficient just means yet more rounds of robbing Peter to pay Paul.
Why should I have to pay more for service, in order to subsidize running bandwidth out to the exurbs and the boonies? If they want high-bandwidth internet connections, they can call DirecWav. If they want high-bandwidth, _low-latency_ internet connections, they can move to someplace where there are a few more people around.
A nit: the U.S. population density is about 27 people per square kilometer, not per square mile. The factor-of-10 ratio, though, seems correct, so I'm guessing the Japan and Korea numbers are also per-km^2.
Comments are Closed.