At last - the first 3G femtocell service has been announced - and it's not exactly as expected.

Femtocells, like all other parts of the technology market, are going to have to adapt to the new financial environment. In fact, as a new technology wanting to roll out into the capital-hungry telecoms world, they may have to adapt more than most - and the news that Softbank is to launch a femto-based service in Japan, in January 2009, shows how this may happen.

We've been waiting for femtocells to reach the market for some time, and this year's Femtocells Europe in London didn't move us much further forward. Lots of secret trials, no public commitments. The Femto Forum defined a standard for "RAN gateway" femtos, that connect to today's 3G networks,and most people thought that these would be the first to arrive in homes.

Softbank has done something slightly different. It's got IMS - the much touted, but so far slowly implemented, IP-on-the-core standard for mobile networks - so it's issuing an IMS femto. This means it's attaching to a flat IP network, not gatewaying into a non-digital network.

It turns out that while 3G handsets work fine on IMS networks, there are data-oriented services that they don't do so well, so IMS networks really want IMS-aware phones. The IMS femto can mediate better data service to existing 3G handsets, so it lets Softbank sell more data services, before all the phones are upgraded to IMS-aware phones.

"It's not a given that a 3G handset is IMS compliant," Keith Day, vice president of marketing at Ubiquisys told me. "If a company is moving ahead on a flat IP architecture, it's a good option to go to femtocells."

What about other femto services? They may be inevitable, but they might still be a little way off.

Back in June, we were waiting for the sub-$100 femto, and wondering if oeprators were going to bother with femtos on 3G networks, when they were so imminently moving to the higher bandwidth and better data services of 4G (LTE or less likely, WiMax).

Now, 4G looks further away, but those operators are still enticing users onto their 3G networks, with data offers, 3G dongles, and Facebook-capable phones. Once these devices start sucking bandwidth, the operators may be staring a network overload in the face - in 2009/10.

At that point, 3G femtos could be the cheapest way to upgrade network coverage and gain extra backhaul (snaffled from users' home broadband).

-----

And as an aside, operators who could go easily to femtos might include those who already have Wi-Fi based FMC services (like Orange's Unik). Often positioned as rivals, these two technologies are just different ways to get 3G traffic off the mobile network and onto the Internet - anyone who has a Wi-Fi based FMC has the makings of a RAN gateway suitable for femtos.