Broadband wireless base stations could get much cheaper, and phones could have a talk-time of days, if a new, long-distance UWB-like wireless technology is for real.

xMax, from xG Technologies, sounds like a contravention of the laws of physics. Yet the company has promised a public demonstration in September, delivering 40 Mbit/s over 15 miles, using less than a Watt of power, at low (sub-GHz) frequencies that have so far been considered unsuitable for broadband.

For now, the credibility of the technology rests on a few web references and the word of a Princeton professor. But xG are doing a better job of explaining their technology than previous companies that have turned out to be snake oil.

"Our unique benefit is we can use narrow channels [at low frequencies] where you get better propagation characteristics," says Joe Bobier, the technology's inventor, and president of xG Technologies. "If you can use these frequencies, yet still be broadband, then you compare favourably with fixed wireless technologies that are doing things in the microwave area where you have line of sight issues."

Magic - or reality?
Sub-GHz frequencies penetrate well, and so can be used for communication without the line-of-sight issues that occur at higher frequencies. However, they are sliced into small narrowband channels by the licensing authorities, so no-one can get hold of enough spectrum to send high-throughput (broadband) data in this spectrum by conventional means. Shannon's Law requires a wider band of spectrum to carry more data.

xMax uses one narrowband channel, but sends more data than could possibly be fitted into that channel - and yet it does not break the laws of physics, claims Bobier. The narrowband channel does not carry the payload, it is just used to co-ordinate and synchronise the sender and receiver. The information is then transmitted in a wideband signal (which xG calls a Flash Signal), at very low power.

"Our technology uses a narrowband channel, and places a carrier there for an extremely precise clock in the receiver," says Bobier. "The transmitter also transmits information in side bands, at levels lower than ultra-wideband. We are able to get performance comparable to a wideband licensed trasmission."

The data is sent out-of-band, but at power levels that are lower than those permitted for unintentional out-of-band signals by normal transmitters. "The levels of out-of-band emissions are highly regulated," says Bobier. "Ours are even lower."

The synchronisation allows the receiver to pick out signals below the noise floor, claims Bobier, so a signal that is not registered by other spectrum users, can be picked up by the intended recipient. There is some explanation in a FAQ at xG, and an article in Microwave Engineering.

Using power levels lower than the out-of-band leakage from normal transmitters makes it somewhat similar to ultra-wideband (UWB - read our coverage), which operates over short distances, using power levels lower than those which are permitted (by the FCC and other regulators) to leak from non-transmitting devices such as stereos.

In September xG promises to have a base station between Fort Lauderdale and Miami, covering a radius of 15 miles, with a shared bandwidth of 40 Mbit/s, all provided by less than one Watt, using the unlicensed 900MHz ISM band. The first demonstration will be omnidirectional, but greater throughout would be possible with a sectorised station, or using a wider noise-level band.

Other benefits
"What Joe has invented can use 100,000 times lower power than other technologies," says xG's chairman Rick Mooers. Besides allowing cheaper base stations for broadband, it has other benefits.

The low-power channel it uses can overlap with other users, because it is below the noise floor, creating "dual use" for the radio spectrum, claims Mooers.

It also has implications for battery life. "Instead of having a 600mW handset, you could have 1mW or 0.5mW handset, says Bobier. This would give a talk time of days. Low power handsets would also neatly sidestep allegations of medical harm from handsets, the pair assert.

The technology can also be used on wireline systems, sending data further over ADSL, and allowing more channels on broadband wired and cable systems, they claim.

Who will use it?
As the name implies, xMax is intended as a "beyond-WiMax" technology, and xG hope to propose it as a potential new physical layer for WiMax. "A great deal of the WiMax standard is essentially protocols, only a little is the physical layer," says Bobier. "This could be super-WiMax."

In the prototype, xG will; use portions of the published WiMax specifications, making this a part-WiMax, pre-XMax network. The name xG is, as you might have guessed, a similar play on the term 3G.

Although currently proposed for WANs, xMax could also be used at LAN and WAN distances "We can go 63 ft with 0.5 pW," says Mooers. "It could run for years on a watch battery." So we can possibly expect x-Fi and and xTooth in the future...

But is it real?
"People will understandably be sceptical as they are with all disruptive technoligies," says Mooers. "We've had five years of rigorous testing, and we say come see it. We are very close to a field demonstration now. That's why we've decided to stick our head up and talk."

The technology is endorsed by Stuart Schwartz, professor of electronic engineering at Princeton University, who also serves as an advisor to xG, and co-wrote the article in Microwave Engineering.

xG will be formally launched in November, and presented at the 3GSM show in Barcelona next Spring. At this point we will start to see the independent testing that will tell us if this is more than a pipedream.