As 802.16 WiMax products finally arrive in the market, vendors have made a belated realisation. The fixed version of the standards, 802.16-2004, which they can implement now, is very different from the mobile version, 802.16e, which is still under development.

Some observers reckon that mobile WiMax will be where the action is, and fixed WiMax will just be a fillip for the existing proprietary fixed wireless market, while others think that success for fixed WiMax might be one of a number of factors that could strangle mobile WiMax at birth.

Either way, the divide is making the whole WiMax scene look less attractive. The standards are separate, and there will not be interoperability between them. Could the WiMax scene wind up like a pantomime horse trying to go in two directions?

"The excitement is WiMax, is not about what's going on at the moment," says Infonetics analyst Richard Webb. "The excitement will be around mobile WiMax."

Gabriel Brown, of Unstrung Insider disagrees: "For all its promise, mobile WiMax will have difficulty finding a market big enough to generate the economies of scale that will permit it to truly thrive.”

Big players vote for mobile
The fixed WiMax market is dominated by Alvarion, using Intel's WiMax chipset, while the rival Fujitsu silicon is in use at Wi-Lan and Terabeam. But whichever dominates, the people with most interest in fixed WiMax are the ones already doing fixed wireless by proprietary means.

"The early part of the market will be proprietary broadband players, who have an opportunity to get established with WiMax products," says Webb. The fixed WiMax specification is stable, and the major players have migration paths.

Several players, such as Motorola, Alcatel and Nokia, have decided to bypass fixed WiMax and wait for mobile WiMax.

"We are striking a slightly cautious note at the moment," says Webb. "We're forecasting gradual growth through 2006 while the fixed standard becomes more widespread. The size will be reasonably low, for a market that will sell into backhaul for wireless service providers and 3G networks."

"Within a year of mobile WiMax products, they will outsell fixed WiMax products," says Webb. "That's where the real excitement about this market will come. In late 2007, there will be 802.16e products, and we will see significant take-off."

Mobile is a moving target
But while the industry holds its breath for mobile WiMax, it may simply be ignoring the problems with that later standard.

"Mobile WiMax is still in the lab, yet is potentially more disruptive and way more exciting," says Gabriel Brown of Unstrung. "It is also an opportunity to leapfrog Fixed WiMax and encroach on the 3G infrastructure market."

However, there are problems. Firstly there's the incompatibility between the two specs. They use different physical layers.

"The fact that carriers won't initially be able to connect to a fixed WiMax base station using a mobile WiMax modem (and vice versa) is a big deal," says Brown's colleague Dan Jones of Unstrung. "It may even hold some carriers back from implementing WiMax, when fixed products become available late 2005 and into 2006, because the carriers would prefer to wait for mobile productions. And that would mean that WiMax rollouts would be delayed until in 2007 or even later."
"Perhaps the biggest barrier to the adoption of mobile WiMax is the lack of available spectrum,” warns Brown, in an Unstrung report, Mobile WiMax: Who Goes Where?. “It’s a tough problem - and one that isn’t likely to be resolved soon.”
The 3.5GHz spectrum often considered suitable for WiMax is only licensed for fixed use, not for mobile usage. While, in the UK, Ofcom is liberalising so that issues like this may be overcome, elsewhere spectrum limitations could be a sticking point.
"Spectrum liberalisation is a proposition fraught with both political and commercial challenges," says Brown. "As a result, it’s hard to envisage near- or medium-term deployments of mobile WiMax outside of the U.S. and Korea.”
In the end, mobile WiMax may be limited to low-mobility applications with a speed of around 0.5 Mbit/s, which could overlap seriously with fixed WiMax. Brown believes that by the time mobile is allowed, vendors of fixed WiMax services may have found ways to offer mobile services.
This could include tweaking the self-install modems and laptop cards, so they can be used in a portable manner - switch on in a new place and start work. Or else, the fixed providers may offer Wi-Fi mesh to fill in the gaps - and use existing client hardware with no modification.

The 3G squeeze
Mobile WiMax will also be squeezed by 3G data, which will could offer similar data rates and coverage by the time mobile WiMax exists.
"With Wi-Fi extending its capabilities, and 3G coming along to handle data, even the mobile part could get snipped in a pincer movement," admits Webb. "Some WiMax services sound a lot like what 3G is all about. In some markets service providers may say 'why do we need 16e?' If service providers don't need a technology, they aren't going to use it."

However, that's not enough to make him write it off: "With that many service providers behind it, you have to assume that this has got legs."