Its probably time to admit it: 3G wireless is like satellite. It will never be economical as a broadband solution.
Five years is a long time to wait and now that 3G wireless is here, youd think wed be grateful. Not a chance of that! We want its successor on schedule, please.
The concept of 3G phone networks was originally seen as a 2001 technology. Then, after four or five years, we were supposed to start using advanced data extensions to 3G.
Well, the news that Flight 123 was delayed is no concern to the passengers on Flight 345; they still expect 345 to be at the gate on time. The worlds desire for wireless broadband didnt disappear just because video phones werent ready for the marketing boys.
So the arrival of both IP Wireless and Flarion on the scene as purveyors of genuine high-speed packet-switched data to mobile users is forcing the phone business to re-think its priorities—especially in Europe.
Its far, far too soon to suggest that either of these newcomer technologies will dominate. Right now, the best bet would be that theyll find separate niches. But the plan was that they would find niches supplying data to people who were, around now, starting to lose patience with wideband CDMA (Code Division Multiple Access) or other 3G networks, and needing more capacity.
“You can see that 3G voice technologies arent suitable for data except on a small scale,” said John Hambidge, global marketing director of IP Wireless, when announcing new licensees of his companys technology. “They are simply not capable of supplying enough data to enough users at a high enough speed to pull prices down.”
That is ironic of course because this was always obviously true about satellite broadband. It doesnt take a very wide calculator to see that satellite data on a penny-per-bit basis cant be made cheap enough to compete in any area where there is actual competition. The technology will always be useful in places where you cant run cable, and people there will always include enough big-spenders who can afford satellite fees.
But satellite will always be a backup, a failsafe, or a last resort. If the launch vehicles and the transponders could be combined for a price of a thousandth of the current budget, then sure! You could send enough of them up to supply 512KB to a large number of homes. But that simply isnt going to happen. When you charge around 500 euros for 50MB of data, your customers are going to have to be desperate, or they wont sign up. And if you have to charge no more than what Europes fiber providers can charge for metro Ethernet, you cant get the rocket off the ground.
Next page: Preparing for takeoff.
Page 2
: Preparing for takeoff”> Well, it turns out that the use of wideband CDMA for data is popular with big spenders. And indeed, personally, I wouldnt be without my Vodafone 3G data card. For someone like myself who absolutely has to have Internet access on a daily basis or go out of business, the price is not the critical factor. At least, not until theres a choice.
The trouble is, the plan was for the 3G network operators to have had four years of revenue at this level before the Flash-OFDM (orthogonal frequency division multiplexing) and TD-CDMA (Time Division-CDMA) packet-switching networks were needed to cope with the overspill. The plan was for the big spenders to have funded the rollout of the network, which would then carry voice to the millions of people who were happy with a 15K bps codec to turn voice into data and back again, but who wanted to send the occasional frame of video.
And of course, the latency is the killer. Both satellite and 3G were conceived by people whose minds thought of “broadcast”: A big antenna owner sending out precious content to an eager audience. If it started a second or so after real time, who could tell?
But data users are running software that was developed for use on local Ethernet systems. They expect communications protocols to run at the very least at around 10M bps with a latency of around 5 milliseconds. Their software times out waiting for satellite or GPRS responses for FTP or HTTP interchanges, and needs rewriting. And instead of being passive recipients of data, they generate it.
So 2004 was the preview for packet-switched mobile data. And 2005 will turn out to be show time. And on this occasion, it may well be that America wont have to regard itself as being a laggard.
In another twelve months, when were preparing our “What happened last year?” summaries, we may well find that commercial deployment of TD-CDMA and Flash-OFDM in North America will have heavily outstripped Europe and even Asian deployments.
Why? Well, the obvious problem with the European and Asian countryside is that its all too easy to dream about do-it-yourself solutions. China is seriously considering rolling out a Wi-Fi franchise system so that users can roam from the nascent mobile phone network onto the Internet, seamlessly switching from one to the other. Britain is planning a network of telematics nodes along all major highways (so are several other European countries) that could provide very high-speed Internet access.
In the wide open spaces of North America, however, the idea of providing a mast every 100 yards along all major highways looks very much less like a commercial proposition. Something with longer reach is essential. And it will have to be a genuine broadband solution, not a slapdash offering that is basically ATM (asynchronous transfer mode) circuit-switched technology.
This time next year, I think the politics of 3G will be in turmoil as the operators try to bargain for better deals, but I think the rollout of widespread high-speed wireless data will be well under way, and reaching close to 1 percent of the data user population.
Read Guy Kewneys other recent columns about trends in mobile and wireless technology.