The Sarasota, Florida-based company, which is only selling less than 9% of its stock in a deal that will value the company at up to $750m, says it can offer a wireless technology that could cover the US for $15m, while Sprint has plans to roll out a WiMAX network for $3bn.

Even worse for the established players, it is offering a disruptive technology in the most disruptive way, helping new companies enter the market using unlicensed spectrum and take market shares from existing players with a far cheaper technology.

If xG lives up to its billing, it will savage the stock market valuations of companies with extensive wireless IP such as Qualcomm and Nokia and allow newcomers to bite huge chunks from the market shares of big carriers such as Vodafone and Sprint Nextel.

But the unspoken question hanging over the one of the most eagerly awaited IPOs in the tech sector since the dot-com boom is quite simply: is it too good to be true?

xG has been frugal in releasing details of how its technology works. It says its patented xMax technology greatly enhances the range and efficiency of communications signals, has a greater signal range and on equal power and bandwidth, can shift significantly more data.

It is based on the discovery by its CTO Joe Bobier of single cycle modulation where individual cycles of RF energy are modified to carry one bit of information, while it says traditional technologies can use tens to hundreds of thousands of waves to convey one bit of information.

Now while the IT industry has been used to new technologies springing from universities or expensively tutored brains in R&D departments, Bobier received his initial training in the US Navy where he studied advanced electronics and satellite communications. He earned the First Class Radio Telephone Operators License, the highest licensing designation recognized by the FCC.

Nonetheless, xG has let the head of Princeton University’s Electronic Engineering Department, Dr Stuart Schwartz into the secrets behind its technology and he has written a report giving a favorable evaluation of xMax. He said that xMax was between 79 and 871 times more efficient than CDMA and GSM.

xG has fended off attempts by existing firms in the industry to have a closer look at the technology, though licensing it to companies with existing channels to market would appear to be a more simple option.

A research note by London brokers Hitchens, Harrison and Co, which are handling the IPO, says the harsh reality behind this is that, especially in the US, patents offer scant protection to anyone who cannot afford to defend them through what are likely to be protracted legal processes.

It quotes the example of the case of NTP Corp and Research in Motion (RIM) where it says RIM effectively violated NTP’s patent on the technology behind the BlackBerry device, but an exhausted NTP finally only got a judgment awarding $612m after years of litigation. While lawyers took a third of this payout it says RIM enjoyed the proceeds of billions.

It says the current strategy behind the strategy is to demonstrate decisively that the xG is in financial control of its own destiny, can independently gain market traction and will have a brand clearly associated in the public’s mind with xG.

This would build a substantial moat around the business’s competitive position, which it would be naïve to believe patent protection alone could provide. Hence, the IPO represents a key step along xG’s strategic path, says the note.

It sees another risk from what will be probable attempts to use political lobbying power to generate regulation restricting the usage of unlicensed spectrum, motivated by the desire to protect high levels of sunk infrastructure investment and not to become obsolete.

Rapid roll-out and clear market traction, presented as a fait accompli, as well as access to international markets, thereby diversifying such regulatory risks, would be the antidotes to this. The business plan is formed precisely to optimize a rapid roll-out and fast time to market, say the brokers.

Hitchens, Harrison dangles the tempting $2.5bn that Ebay paid for Skype, with a possible earn-out that could take it up to $4.1bn, as a precedent. It says there will be strong interest from industry incumbents, whose survival in threatened, and all those companies, which have designs on the various communications’ channels, such as Microsoft. It says no company exposed to the space will be able to ignore the technology. This will almost certainly translate into takeover or strategic partnership interest.

It is a range of 7.55 miles that give xG base stations the big advantage of alternative technologies and this far fewer would be needed. It claims to build mobile networks for as low as 4% of rival infrastructure costs to obtain equivalent coverage. XG says it had so far received $20m in letters of interest to secure rights to use its base stations in territories with a population of 20 million in southern states of the US.

It says it can yields the same range for significantly reduced power and this will lead to much greater battery life for devices, such as laptops and phones, that use the technology.

An initial order of 10,000 handsets has been placed with an OEM and the technology is due to be rolled out in the second quarter of 2007 and will initially be restricted to voice, though the data and video requirements of 4G are in the pipeline.

XG’s aim is to concentrate on mobile VoIP initially but eventually aims to sell its xMax chips for a wide variety of markets and it can even see wireline applications.

Anticipating a sceptical response from some potential investors, Hitchens, Harrison uses an equation to show that if, for example, investors think that a company has a 5% chance of making them 50-fold their money, and a 95% chance of losing them all their money, it is still a rational bet in a portfolio.

While investors can choose whether or not to gamble on the xG’s success, existing players in the industry have no choice but to face huge losses if it proves a success. The Europe Commission has estimated, for example, that 3G build out on the continent will cost 260bn euros ($331.9bn), half for licenses and half for infrastructure.

For the industry as a whole the pathway to 4G appeared to be predictable, with the acronyms OFDM and MIMI pointing the way. Now a company with just 23 employees is suggesting a different route.

The nightmare for companies that have invested billions of dollars on communication infrastructure or wireless IP is that xG may not be the last of companies to emerge from garages with technologies that may totally disrupt their existing plans.