Too good to be true?
SUNRISE, Fla. — Wireless communications traditionally has been a story of tradeoffs, with each solution seemingly creating another problem.
Want greater signal range? Increasing the signal power will do the trick, but doing so increases the noise floor, increases interference risk and can drain the battery life from a receiver. Using smaller cells mitigates these negative effects, but that requires more base stations, which can greatly increase capital-expenditure and maintenance costs.
And if you want nationwide coverage, be ready to write checks for billions to pay for the spectrum at auction before spending a dime on network infrastructure.
But those truisms may change drastically in the near future, thanks to the work done in the lab of xG Technology. Within a modest 3500-square-foot space in a Ft. Lauderdale-area strip center, xG President of Operations Joe Bobier — inventor of the company's technology platform — has led a small staff in the development of a broadband-wireless system known as xMax, which uses a patented technique that Bobier characterizes as “a fundamental paradigm shift in the way radio signals are modulated and demodulated.”
Paradigm shifts generally are overblown. But xG promises a quantum-leap package: easily deployable broadband data in a potentially wide coverage area using little or no dedicated spectrum and operating at incredibly low power levels.
“It is a rather unusual way to do communications. In my 40 years in this industry, I've never seen anything like it,” said Stuart Schwartz, a Princeton electrical engineering professor. “I think Joe Bobier is a very, very clever guy. He's done something quite remarkable.”
Ali Hedayati, a former Intel general manager who headed the chip giant's Centrino effort and now is a technical adviser for a venture-capital firm, echoes this sentiment.
“Most disruptive technologies change the way a segment of the business is run,” Hedayati said. “I think xG's technology could potentially do that.”
HOW DOES IT WORK?
At the heart of the xMax solution is xG's Flash Signal technology, which utilizes single-cycle waveforms to transmit information at a minimum effective rate of 1 MB/s for each megahertz of spectrum utilized in the information-bearing channels, Bobier said. Whereas most communications systems require a receiver to modulate thousands, or millions, of radio frequency (RF) cycles for each bit of information, the Flash Signal scheme accomplishes the feat in just one RF cycle.
Not only does this single-cycle technique increase throughput efficiency, it means even traditionally weak signals are usable, Bobier said. Moreover, because the receiver — the design of which is xG's most-guarded intellectual property — includes a passive wavelet path filter that acknowledges only single-cycle waveforms, all other RF signals are ignored. This lets xMax operate in “silent spectrum” that permits effective communication even at very low power levels, he said. (For a more technical explanation of how this technology works, visit "Breaking free of the past"
“We're talking about a 25 to 45 decibel advantage in an industry where 2 decibels is worth killing for,” Bobier said. “And you don't have to throw a lot of spectrum or power at [producing a broadband signal], if you can get rid of the noise.”
By virtually eliminating the typical noise floor, xMax enables the power levels for its information-bearing signal to be as much as 100,000 times below the FCC's current regulated power limit for out-of-band emissions, which is designed to prevent systems from interfering with each other. In fact, xMax's power levels are 10,000 times below the FCC's power limits for ultrawideband (UWB) transmissions, according to xG officials.
IS IT REAL?
Not surprisingly, the notion that an unknown company could devise a communications scheme that will outperform WiMAX at a fraction of the infrastructure costs while operating at sub-UWB power levels is enough to make those in the industry raise their eyebrows. The fact that xG claims it can do this without an operator having dedicated spectrum has caused many to incredulously dismiss xMax before seriously considering the technology.
“Broadband without spectrum — that's pretty funny,” said one industry source upon initially hearing a description of xMax. “Did they also promise you the Brooklyn Bridge when they were selling you that?”
Bobier said he understands the skepticism, particularly among longtime RF engineers who don't understand how the xG Flash Signal technology can possibly work. Indeed, there is no way xMax would have been developed within the constraints of traditional RF thinking, he said.
“There's nothing intuitive about processing radio and power in single cycles,” Bobier said. “Every time I found myself frustrated and I was going wrong, it was because I was following all the training I had received. It was only when you have a moment of clarity and you tell yourself, ‘Stop thinking in terms of power and tones and start thinking in terms of the mechanics of one wave at time’ that you make progress.”
Robert Syputa, a senior analyst at Maravedis, said the doubts about xG's technology are largely a reflection of the company's large promises.
“It's never proper to suspend suspicion about something that portends to walk on water,” Syputa said.
Indeed, Bobier encountered initial resistance even within xG's hierarchy when he first broached the notion of single-cycle modulation.
“Initially, we probably had the same reaction people are having today upon hearing about it for the first time,” said Roger Branton, xG's chief operating officer and chief financial officer. “It just sounds too good to be true.”
But xG officials insist it is true, the product of years of development and a few key moments of inspiration (see sidebar). And, for all the questions raised by the engineering community about xG's modulation scheme, there are an increasing number of indicators that xMax is more than an empty promise.
Princeton's Schwartz, an admitted initial skeptic, today may be xG's staunchest advocate. Hedayati said he believes xG's technology “has legs” after following the company and its technology for months and seeing a lab demonstration in August.
Perhaps most significantly, SkyTel — a paging firm that is a wholly owned subsidiary of MCI — has validated xG's technology and signed a letter of intent with the fledgling company in June, said Rick Mooers, xG chairman and CEO.
SkyTel declined to be interviewed for this story, but its belief in xG's ability to revitalize sub-gigahertz spectrum caused the paging firm to scrap plans to sell its 900 MHz airwaves, Branton said.
“They totally changed their business strategy after seeing this technology,” he said.
Of course, lab demonstrations can only be so convincing, which is why xG plans to conduct an xMax trial during the next two months in preparation for a scheduled commercial launch of a broadband wireless offering in November. The fixed-wireless system will have the following attributes:
Antenna: One omnidirectional transmitter located on TV broadcast tower at a height of 850 feet;
Range: 20 miles, covering the cities of Miami and Fort Lauderdale, including inside buildings (see map);
Data rate: At least 10 Mb/s throughout the coverage area;
Spectrum used: A 6 kHz carrier on dedicated spectrum and 10 MHz of information-bearing spread spectrum in the unlicensed 900 MHz band;
Interference: None, despite the presence of a TV transmitter on the same tower;
Power: A maximum of 50 W for the carrier and less than 0.15 W for the information-bearing channels.
The success level of this trial will largely determine xG's fate, Schwartz said.
“I understand why it works and how it works; what I don't know is just how well it's going to work when you place protocols and applications on top of it,” he said. “There may be some loss in the data rate … but I don't see any fundamental, technical reason why this won't work in a multi-user environment.”
WHAT DOES IT MEAN?
If xG's technology works as promised, the potential ramifications are mind-boggling, by all accounts.
As a fixed-wireless play, it could provide a local-loop alternative for independent ISPs and competitive carriers, which effectively have lost all leverage in their negotiations to lease lines with cable operators and telecom carriers that provide access to their customers. Mooers said SkyTel's representatives noted the local-loop potential of xG's technology during their Florida visit.
“If this technology had been available four years ago, AT&T and MCI probably wouldn't have been in play,” Mooers said, noting the long-distance giants' inability to connect directly with customers forced them to merge with RBOCs.
For operators with existing spectrum, xMax offers an opportunity to re-use the airwaves and maximize the value of the asset. This could be particularly significant to public-safety entities. Instead of deploying hundreds of nodes at 2.4 GHz or 4.9 GHz for a few square miles, a public-safety organization could deploy xMax to provide a fixed-wireless connection for a much greater coverage area with just one base station while re-using existing sub-gigahertz spectrum.
For potential operators without spectrum holdings, xMax can let them enter the wireless game. Instead of using dedicated spectrum, the timing function can be integrated with the information-bearing channels at a loss of 25% of the data rate, Bobier said.
Hedayati said he believes xMax could spark regulatory changes, including the way the FCC allocates spectrum and its interference rules. But the more immediate impact will be the manner in which it changes the economics of entry into the wireless market, he said.
“You can actually deploy [xMax] on unlicensed spectrum — that becomes a completely different ballgame,” he said.
One of the most intriguing aspects of xMax is its potential in the mobile wireless environment, although the technology has not been optimized for that use yet. Because xMax operates at such low power levels, the battery life of mobile devices could improve dramatically, Bobier said.
The low power levels also create other opportunities, including Zigbee-like devices that offer greater range and functionality. For law enforcement and the military, the low power levels would make it easier to communicate in “stealth” mode because only an xMax receiver can detect xMax transmissions.
And xG Flash Signal technology is not limited to the wireless arena. Using the single-cycle waveform and receiver-centric design, telecom carriers should be able to quadruple the range of their DSL offerings, and cable operators can double the capacity of their networks, Bobier said. Similar advantages should be achievable in optical technologies, although xG has not tested those possibilities yet, he said.
“It's what we want it to do,” Bobier said. “We're starting with the bricks, and we can build any building that we want.”
That characteristic makes xG's technology especially attractive, Hedayati said.
“The beauty of this thing is that you don't have to modify anything — you can just walk right into existing infrastructure,” he said. “You can be on licensed spectrum or unlicensed spectrum. … You can have more flexibility with this than anything I've ever seen because of the way they deal with the noise.”
Officials for xG acknowledge all these possibilities and say new applications for the technology emerge on a weekly basis in different markets. But don't expect xMax to fund trials for anything other than the fixed-wireless project, Mooers said.
“If it's viable [as a fixed-wireless product], the engineers will see that it works, and they can extrapolate the application to other markets,” he said.
WHICH WAY TO THE PROMISED LAND?
This philosophy underscores a fundamental reality for xG: Its technology cannot reach its full potential if the small company tries to do everything itself, Mooers said.
“We're going to do fixed wireless in a Cisco model, where we make high margins on the equipment in the short term,” he said, noting that some wireless ISPs may let xG Technology get a piece of their revenues. “But it shifts to a Qualcomm and royalty model when you hit the economies of scale and large volumes. It will end up being small amounts of money — but in billions of units — if we hit it in the right scenario.”
That means teaming with companies that bring assets to the table outside of xG's core competency, Mooers said. The list of potential candidates is huge — chip vendors, independent ISPs, software companies and service providers, to name a few — and the most attractive possibilities are massive companies because they can provide the financial, legal and political resources that can assure xG's single-cycle system makes it to market on the company's terms.
“We're sitting on the solution,” Mooers said. “A lot of people say, ‘You're a competitor of WiMAX.’ That's not necessarily true — we may very well be a key part of the WiMAX solution. … It depends on who our partner is. If our partner is Intel, it's more than likely going to be kept under the name WiMAX.”
Hedayati said he believes the potential for such a scenario increased greatly with Qualcomm's recently announced plans to purchase OFDM vendor Flarion Technologies (read "OFDM sector faces changes"). With most WiMAX technologies not ready for prime time, wireless operators looking to deploy 3G and 4G technologies have little choice except to pay royalties to Qualcomm unless they turn to xMax, he said.
But striking such agreements promises to be a delicate endeavor, Mooers said. To date, most inquiries from large companies have come from their venture-capital arms — a direction in which Mooers is hesitant to go, claiming that funding is not an issue for xG. He said he is not interested in offers that would let a large company effectively swallow xG and its technology.
Mooers also said xG will be very discriminating about any exclusivity arrangements, probably granting exclusivity only on a niche-market or geographic basis, if at all.
“We have to make sure that we hook into a revenue model that lets us get a good piece of the economic value we provide,” he said. “We're not just going to do another big company/small company deal, where the big company gets the lion's share of the economic value from the small company's innovation.”
Complicating matters is the fact that many potential partners may have mixed feelings about xG's technology. For example, cable operators might view xG as an opportunity to add a wireless component cost-effectively to their portfolios or they might see it as a threat to their last-mile holdings.
Similarly, some wireless carriers may be excited that xG will let them add new services through the re-use of their spectrum, while others may curse the fact that this new technology could negate the current spectral advantages for which they've spent tens of billions of dollars during FCC auctions to secure.
Even the government could have reason to be ambivalent. Certainly xG's technology promises to be a great way to make President George W. Bush's goal of ubiquitous broadband in a competitive marketplace a reality. On the other hand, Congress is counting on a 700 MHz auction this decade to generate at least $10 billion to help address budget deficits, but bidders might not be willing to pay that much if a spectrally efficient technology like xMax lets them compete effectively without owning a lot of airwaves.
Of course, all of these hypothetical scenarios will be irrelevant if xG does not convince the marketplace that its technology is for real. xG officials say the stated performance projections for the technology are conservative, but Mooers recognizes that only a successful deployment will completely convince the naysayers.
“After this trial, I think [potential strategic partners] are going to come to us, if only because they risk having to compete against us if they don't.”
INSPIRATION HELPS FILL A NEED
It often is said that necessity is the mother of invention, and that certainly was the case with xMax. And, if Joe Bobier's invention of xMax somehow proves to be troublesome to Verizon Communications, the giant telecom carrier would have only itself to blame.
That's because Bell Atlantic — the RBOC's name before its merger with NYNEX — provided the copper wires that Bobier's family-owned business in Parkersburg, W.V., depended on to support its dial-up ISP business. But the RBOC wires frequently failed to provide a good connection, causing Bobier to believe that an alternative access method to his customers was necessary.
“I ended up with a staff of something like nine people working about 18 hours a day answering tech support because of phone-line problems,” he said. “When I looked at the cost of the phone lines and the cost of supporting the poor quality of the phone lines, I realized that's where all my money was going. I needed a way around.”
Thus, long before the term Wi-Fi was coined, Bobier developed several broadband wireless technologies in the 2.4 GHz band that was the foundation of a successful business.
“We built a very effective citywide network and started selling broadband service in 1997. I was selling 1.5 Mb/s speeds for $29.95,” he said. “I couldn't put it up fast enough. I covered my town, the neighboring town and then two more towns.”
While the wireless link was better than Bell Atlantic's copper lines, it was far from perfect — access points cost almost 10 times more than they do today, making the economics of a 2.4 GHz network even more challenging. Instead of dealing with multiple access points, Bobier wanted to cover a town with a single base station, like the systems he worked with in the Navy and as a land-mobile-radio dealer.
One snowy night in November 1999, the answer to the problem came to Bobier during one of his frequent midnight reading sessions in his country log house in St. Mary's, W.V. While reading Stephen Hawkings' “A Brief History of Time” and musing about Albert Einstein's Nobel Prize for explaining the Edison Effect, Bobier had a revelation.
“It just occurred to me that the least amount of radio energy I could have is one photon — one wavelength — and the least amount of data I could have is one bit,” Bobier said. “It just struck me that there was a correlation there.”
Bobier said the notion so inspired him that he began doodling and taking notes furiously for more than an hour: “I was so high then, there was no way I was going to bed.”
Filled with inventive adrenalin, Bobier decided to put his idea to its first litmus test by awaking his wife, Angie, during the early-morning hours.
“She's not a technical person at all, but she's smart,” he said. “I drew this correlation for her between one particle of light and one bit of information. I explained this whole idea, and she started asking questions. Finally, she said, ‘Yeah, I think you might have something.’”
Bobier decided to take his single-cycle radio theory to Rick Mooers, who had bought his wireless broadband ISP a couple of months earlier.
“I brought the idea to [Mooers] and told him that it was a long shot,” Bobier said. “There was a lot I didn't know yet, but if my intuition was right and my understanding was right, this would potentially be a groundshaking thing.”
Mooers agreed to fund six months of initial testing, which was successful enough to convince Bobier that a single-cycle radio system could be built. That prompted additional funding from Mooers and the establishment of a lab in Ft. Lauderdale, Fla.
By fall 2003, the xG Technology team had resolved most aspects of the system but continued to get mediocre-to-disappointing performance from all its receiver designs — a fact that a technical consultant to xG Technology's then-CEO Mats Wennberg noted repeatedly during Bobier's visit to Stockholm, Sweden.
After a day full of “razzing” by the consultant, Bobier was admittedly rattled about how he would get the receiver to work correctly. He returned to this “crummy little European hotel,” only to find that a stuck window was letting all the cold night air in his room. Frustrated and cold, Bobier laid on his bed thinking about the receiver when he was hit by another piece of midnight inspiration.
“I stopped thinking in normal terms, slowed everything down and started visualizing the waveform in real time,” Bobier said of the patent-pending design. “All of a sudden, it came to me exactly what I needed to do. … It's so incredibly simple.”
Too simple, in fact. When Bobier explained the concept to chief engineer Nadeem Khan, the notion was not met with enthusiasm.
“So I just went back to work at my desk for a week and watched him struggle [with receiver designs], making all the same mistakes I had been making,” Bobier said. “Finally, I turned to him and said, ‘Are you ready to simulate my circuit?’
“He said, ‘Yes,’ and went to work. It only took us 15 minutes to simulate, and the result was dead-on perfect. And that's the receiver we use today, with virtually no changes at all.