The broadband fixed wireless industry, much like its wireline brethren, fell victim to its own hype...and it hasn't been a pretty picture. The fixed wireless industry seemed to nosedive following the much-ballyhooed spectrum auctions that occurred in the mid- to late 1990s. Billions of dollars were spent to snatch up available spectrum with MMDS (Multichannel Multipoint Distribution Service–2.1 GHz to 2.7 GHz) licenses and LMDS (Local Multipoint Distribution Service–27.5 GHz to 29.5 GHz) licenses.

Three years ago, companies like Sprint and WorldCom spent nearly $2 billion each to snag roughly 30 percent of the nation's MMDS licenses. Meanwhile, other companies like Teligent and Winstar plummeted into bankruptcy after spending billions of venture capital dollars to build out markets across the country.

To add insult to injury, first-generation fixed wireless equipment failed to do the job, says Russell Wiseman, senior vice president– Internet operations for Nucentrix Broadband Networks (www.nucentrix.com), a fixed wireless service provider that serves nearly 100 small to medium-sized markets across Texas, Oklahoma and the Midwest.

Fixed wireless' evolution

"The industry," says Wiseman, "has come through an evolution working with the only commercial systems that were available (at the time), learning the lessons, figuring out what worked and what didn't work, and what was required to make the business model all fit together.

Wiseman

"We came to the conclusion that the first-generation systems, which were pure line-of-sight (LOS) systems with external antennas, were not going to make the business model work. From that moment, (we) started looking to second-generation platforms to make the business model work."

Wiseman has a clear idea of what second-generation equipment should be. "Second-generation, for me," explains Wiseman, "is non-line-of-sight (NLOS).

"When I refer to 2G, I refer to the system that makes the complete leap to an NLOS system with a cheap and inexpensive CPE that the subscriber would implement themselves."

Taking a cellular approach

The failure of the first-generation "single stick approach" that featured high-power, large external antennas that couldn't penetrate any kind of foliage or obstruction, meant the base station had to be brought closer to the CPE, says Wiseman. That means a cellular approach.

But, with a cellular approach, vendors are faced with what Wiseman calls the Rubic's Cube issues.

"I want to be able to have high-speed data, because if I don't, I don't have a product," says Wiseman. "But, in order to deliver that high speed, I don't want to reduce the cell radius so much that I have to raise more money than planned (to deploy more cells). I don't want to make the CPE so big and expensive that the subscriber can't install it, or that I, as the carrier, can't meet my cost-of-acquisition targets because the CPE is too expensive. Those three issues are the engineering challenges of all the vendors."

Faced with these challenges, the vendors are taking various approaches to solve "those dilemmas, those fundamental inconsistencies or conflicts in the business model," Wiseman says. They're applying techniques like OFDM and CDMA. They're debating modulation schemes like 64 QAM and QPSK, as well as discussing whether to pursue a mesh architecture or a large cell network design.

Various company representatives expelled a great deal of air on whether OFDM or CDMA was best. For Wiseman, it's an irrelevant debate because there's always going to be benefits and tradeoffs, no matter which access scheme is used.

He says that OFDM, while a relatively new approach, is very robust in the face of multipath and other interference. While CDMA is not as strong in that regard, it is a more established approach, with nearly 10 years of development behind it. As a result, this could translate into cheaper CPE using the more mature technology.

But Wiseman doesn't believe the OFDM/CDMA, 64 QAM/QPSK or the mesh/large cell debates will define who wins or loses in this space. Instead, the one constant that's having an impact on everyone in the industry, vendor and service provider alike, is time, he says.

The clock's ticking

Wiseman explains: "It comes down to which vendor is able to mature their technology the fastest to allow us, as operators, to launch networks before the window of opportunity closes for us and them.

"If you believe that there is a window of opportunity for carriers like us (i.e., Nucentrix, Sprint, WorldCom); if you believe that the market opportunity will not be there forever, that DSL and cable modems will continue to make inroads as every day goes by, the window for a third alternative becomes smaller and smaller.

"So, we can't wait around forever. We have budgets. We have cash balances. We have investor impatience. All those market opportunities or lack thereof have to be addressed in getting to market."

How much time do the fixed wireless providers have?

Wiseman says that in 2002 we'll start to see some new generation equipment actually take form and make it to trial. 2003, he says, may be the crucial year for fixed wireless.

"That's when momentum needs to be built, or people are going to lose interest. Within that window of time, a good 15 months or more, a lot is expected."

Wiseman realizes his company's decision will not make or break any one vendor. But, he says Nucentric's decision will not be made in a vacuum, either. "Our strategy," says Wiseman, "is to make a decision partly due to our own assessment of our own needs, and partly based on what bigger carriers do.

"That's borne out of the fact that we're a big operator, but we're not a BIG operator. We're bigger than guys who have only one license in one market. But, we alone can't drive the volumes of a vendor to sustain them on our own. We need large volumes from a WorldCom, from a Sprint, or a Verizon to help drive healthy vendor industries.

"So, we're preparing, based on our own evaluations, to make a choice, and we're watching what the bigger vendors do so that we're not caught off guard and find ourselves with a technology that nobody else is deploying."

However, Wiseman believes, despite the Rubic's Cube challenges vendors face and the fast-approaching window of opportunity, there's hope.

"If you look at where the technologies are today," says Wiseman, "where they were six months ago, and where I expect them to be in six months' time, I'm pretty confident we're going to have a couple of potentially strong contenders to pick from."

Wholesale fixed wireless?

By all accounts (literally and figuratively), Dean Johnson, president and CEO for First Avenue Networks (www.firstavenet.com), should be wary to the point of being frightened about the transition the fixed wireless industry is going through. That's because, in its previous life, before the bankruptcy, First Avenue was ART (Advanced Radio Telecom).

Johnson

First Avenue Networks emerged when ART came out of bankruptcy in late December 2001. In its new form, with a new management team and board of directors, it has nationwide coverage of 39 GHz licenses with deep coverage in the top 20 United States markets. Its total U.S. license portfolio of 753 licenses represents 904 million channel POPs.

Despite the bankruptcy backdrop the industry has, Johnson says there's still a positive side to things. "That's the environment we find ourselves contrasted against," says Johnson. "But, here's the good news in all this. There's been no change to the business fundamentals.

"We see the demand and the opportunity is just as great as ever. There's really been no diminishment of that. We're getting second-generation technology now. It's technology that was targeted for the Winstars and the Teligents of the world that went into the labs two years ago and is finally coming out. I've seen the features to be on target and the costs are down."

Mapping out the future

Given that, what's a formerly bankrupt company to do? Johnson has a clear path mapped out. "What we're trying to do," says Johnson, "is put our spectrum to work. We are going to sweat these assets, and that means we're going to focus on market opportunities instead of just building markets and links."

First Avenue, says Johnson, is "going to identify demand and serve it where we find it." What the company won't do, he says, is go after the retail market.

"While going after the retail customer may be a good proposition," says Johnson, "it's very expensive. There are only so many (telecom) companies in a retail market who can afford to go after the retail customer.

"So, what we choose to do is serve those companies that are already serving the retail customer. You can think of it as sort of a wholesale strategy. So, we are looking for large, diversified telecommunication providers that wish to avail themselves of the use of wireless frequencies. We'll even build facilities for them."

Johnson is also encouraged by the technological advancements in fixed wireless. "Gen 1 is ready to go," he says. "The problem is that it's got gen 1 performance and gen 1 cost. Gen 2, I'm very, very excited about. You look at some of the point-to-multipoint developments coming out, and the next generation of point-to-point equipment, that is really getting where you need it to be. That's the good news.

"Knowing that and if I had the right opportunity, I wouldn't hesitate to deploy a competent company's gen 2 equipment. That's because, more so than ever, they have to make that deployment work or they're out of business."

Johnson also believes fixed wireless success in the future is going to be based on strategic partnerships. "We believe that the future solution is going to have to be a collaborative effort," says Johnson, "and it's not going to be just one medium. For example, you need to pair up what we've got with a metro fiber provider. Nobody needs a backhaul provider because there's so much capacity in the Global Crossings of the world.

"We need to make sure that when we do roll out, we have to do it smartly, and we do it using other people's assets while they're using our assets. We may use their sales force while they may use our technical strengths. That's the thing. No one company can do it all."

(Editor's note: During the 2002 Broadband Wireless World Forum this past February, there was a lot of activity and announcements from both established and start-up vendor companies touting their next-generation products. After having gone through a very public bankruptcy bloodletting of some key providers, the industry's remaining service providers, while appropriately chastened by past events, were generally optimistic about the industry's chances for a rebound. CED contacted two such providers who, out of sheer self-interest, are closely following the technological events that are certainly reshaping their industry...and most likely their fortunes.)

Boingo follows the bouncing traveler with national WiFi access For the seasoned business traveler, time spent waiting in public places is usually time wasted. Thant's why when wireless Internet accessibility first appeared (WiFi–the 802.11b wireless standard was approved in November 1999), many travelers thought they had died and gone to the Internet cloud. Travelers swooned at the thought of having wireless high-speed access to e-mail, Web sites and even online discussions or conferences. Unfortunately, like many other touted cyber-related products, the reality of using such services has proved to be more frustrating than fulfilling. Enter the Internet's own wunderkind, Sky Dayton, founder and chairman of Earthlink, the nation's second largest Internet service provider. Dayton launched Boingo Wireless (www.boingo.com) this past December to address the mess that mobile WiFi has become. Speaking to the Broadband Wireless World Forum last month, Dayton tried to encourage service providers to get onboard the WiFi rocket before it reaches "escape velocity." The problem with current WiFi access, said Dayton, is a difficult, if not downright unpleasant, customer experience. It usually involves clumsy set ups with different PC cards or software that has to be used at different access points. Because there are multiple WiFi operators, that means travelers must have multiple WiFi access accounts. Customer service among the hundreds of WiFi networks, said Dayton, is virtually non-existent. To solve the conundrum, Boingo partners with all those (independent) networks and stitches together a heterogeneous amalgamation or patchwork of different networks into a single, seamless experience for the end user. The key, says Chris Gunning, director of product management at Boingo, is to bring all these growing, but separate, networks together in a single offering. This, in turn, enables the creation of "a much more ubiquitous, persistent access to those networks." Boingo looks to aggregate not only large networks, but regional and local networks as well. It is even interested in independent shops, stores or public facilities like airports, libraries, hotels, or office building lobbies. The opportunities, says Gunning, are everywhere. "We see growth coming from all sides. We see it coming from major national providers. Because the cost of the individual location is relatively small (to become a WiFi access point), we also see it coming from the grass roots level. We see this growth on the grassroots side to be on the same order of magnitude in terms of the ultimate number of independent locations as we see on the larger carrier class (type service providers)." Gunning says the growth potential for independent retail locations is particularly important. "Most of the coffee shops and other retail outlets that are being deployed," says Gunning, "have either a cable or DSL line coming in. At the end of that is a little box that serves as a gateway, firewall, NAT and DCP server, with an access point attached to it. "In one of our locations, our software recognizes whose network it is and knows how to authenticate against that gateway so that it will open a firewall rule for you so it will give you access to the Internet." What does Boingo offer? The access locations get national marketing support. Their customers get easy accessibility to Boingo "hot spots" around the nation with Boingo's free client software that configures computers for any Boingo location automatically. They also get 24-hour technical support and customer service. And lo and behold, customers get just one bill. (See chart for service charges.) As a result, says Gunning, "It would be fair to say that we are talking with everyone we know who has either deployed commercial hot spots or is considering deploying commercial hot spots," whether they're independent businesses or public facilities, or they're wireline or wireless service providers. Both Dayton and Gunning believe carriers or established service providers must act soon if they want to get a piece of the WiFi market. They're both of the same mind on WiFi's future–that in the next five years, every high-traffic public location will be a WiFi hot spot. Says Gunning: "It's still early enough that there's a huge land grab going on to install more (WiFi) locations and build out networks (to serve these locations). They have huge networks that are already installed, and they have fixed costs associated with those networks that they have to support. We're putting people into that pipe and generating revenue for them without them having to encumber the costs of customer acquisition, marketing or support." Boingo's cost structure Boingo Pro–Goes for $24.95 a month and includes 10 Connect Days. A Connect Day includes unlimited access in a Boingo location for up to 24 hours. Customers can even disconnect and reconnect within each 24-hour period from the same location with no additional charge. Each additional Connect Day is $4.95. Boingo Unlimited–Unlimited monthly usage for $74.95. Customers can connect all day, everyday. Boingo As-You-Go–No monthly fees, pay for what you use. $7.95 per Connect Day (see above). Fixed wireless standards: One down, one to go After a two-year effort, the global IEEE 802.16 Wireless MAN air interface standard (www.wirelessMAN.org) is the first broadband wireless access standard to be developed and released by an accredited standards body. But like all standards efforts, there's more to come, says Brian Kiernan, vice chair for the 802.16 Working Group and senior vice president at InterDigital Communications Corp. "The base standard, IEEE 802.16, is published," explains Kiernan. "It covers 10 GHz to 66 GHz wireless networks. The first amendment to it, 802.16a, which covers 2 GHz to 11 GHz wireless networks, is in comment cycle right now. There will probably be one more circulation of that before we take the document and recommend it go out for sponsor ballot, which is the full-blown IEEE-level approval. That will happen no later than May." Proponents of the standard believe the approval of the base document sets the stage for the widespread deployment of wireless metropolitan area networks as an economical method of high-speed, "last-mile" connections to public networks. Kiernan says one of the biggest accomplishments the standard achieves is a very flexible MAC (medium access control) layer. "We specifically held up the base document," he says, "so that we could make some changes in its MAC that would allow it to be much more amenable to the coming amendment (802.16a), which is more the MMDS document to provide for really a very, very flexible MAC (see table above). "With that, we can very easily accommodate variable data rates, variable users, different kinds of services, and different kinds of requirements for different users, etc." The flexible MAC, says Kiernan, was developed so that they wouldn't have to completely re-write the upcoming 802.16a document from scratch. That flexibility includes having either single- or multiple-use subscribers. And, based upon the traffic requirements or demands that different users are giving it, an 802.16 unit can make basic decisions with regards to how much bandwidth it's going to ask for from the base station. While he typifies the spectrum covered by the 802.16 base document as "big pipes to big users," Kiernan says one of the challenges with developing the 802.16a document has to do with the markets that spectrum deals with–residential and SOHO markets primarily, or as he characterizes it, "smaller pipes to a much more dispersed population." He says one of the big battles at the very beginning of the 802.16a work was the question of how granular to get in terms of service. Kiernan says those involved finally agreed that what they were trying to do was to compete with cable and DSL modems. As a result, says Kiernan, "the way the MAC is put together and the amendments we've made to the MAC, you can go to a much finer grain. As opposed to dumping tens of megabits to one location, you're typically going to be dumping a megabit or so amongst a hundred or more locations." The 802.16a deliberations were complicated by the different regulatory environments around the world in which the standard must be able to operate. "We have to deal with a lot more variability in terms of the RF channel," says Kiernan. "How narrow is it? How wide is it? "In some cases, you can be talking one-and-a-half megahertz wide channels. In other cases, you're talking 10 MHz or 20 MHz wide channels, which is a big difference. We wanted to try and write a standard that was as 'universal' as possible, so that it was usable in these different regulatory environments." The 802.16 Working Group is also working on a coexistence or recommended practice for each spectrum group, says Kiernan. "Within 802.16," says Kiernan, "we published a recommended practice, which is essentially information to potential users of this standard to say this is how you can use it in various frequency bands. "It's a recommended practice for 10 GHz to 66 GHz and it talks about such things as co-channel interference, adjacent channel interference, deployments, site locations, antennas and things of that sort. We are doing a similar thing for the 2 GHz to 10 GHz." The recommended practice for the 10 GHz to 66 GHz spectrum has already been published. A similar effort for the 2-11 GHz spectrum is currently in development and is slated to be completed by March 2003. 802.16 MAC overview Supports difficult user environments–high bandwidth on-demand; hundreds of users per channel; continuous and bursty traffic; very efficient use of spectrum. Protocol-independent core (ATM, IP, Ethernet, etc.)–ATP-based and packet-based convergence layers. Balances between stability of contentiousness and efficiency of contention-based operation. Flexible QoS offerings–CBR, rt-VBR, BE with granularity within classes. Solid privacy and encryption. Many options for vendor innovation and differentiation–e.g., scheduling. Built to support multiple PHYs.