Fixed: Broken? The future of fixed-wireless broadband.
Margot Suydam, Technology Editor -- CommVerge, 3/1/2002
It seems only yesterday that the telecommunications industry was singing the praises of fixed-wireless broadband access. Service providers both large and small saw it as a fast and easy way to leapfrog last-mile deployment obstacles, not to mention the draconian control of the local loop by incumbent phone carriers. To many hopefuls, the technology looked like another path to high-speed Internet nirvana.
Instead, fixed wireless led to purgatory. First-generation products proved less mature than had been hoped. As the economy went bad, the heavyweight service providers who had planned massive rollouts abandoned their plans. A communications technology that had been hyped as the "third pipe" into the broadband home (after cable and DSL) ended up labeled as a disappointment.
However, proponents maintain that fixed wireless ain't broke. The technology has found some success in rural and international markets. Second-generation systems are starting to emerge, bringing the promise of better coverage and easy installation. And veteran and upstart companies alike see a bright future for the technology.
What befell this market, which was fueled by such high hopes and high investments? Obviously, these have been tough times for the broadband industry as a whole. Fixed-wireless efforts got hit by high infrastructure expenditures and installation challenges in a time when capital funding was quickly becoming scarce.
According to some, the immaturity of the technology itself also played a part. Many insiders now agree that so-called first-generation systems were too limited in where and how they could deliver coverage, which in turn limited their ability to support a viable business proposition for service providers.
As it stands now, vendors of first-generation products face a conundrum: upgrade their technology to play in the big leagues or be satisfied serving a niche in the US. On the other hand, these companies have found opportunities in international markets, where the lack of wired infrastructure makes wireless broadband an easier sell.
Meanwhile, a whole lineup of startups is entering the market with technology approaches that supposedly address key technical and market issues. The hot-button technical issues are self-installation and the ability of fixed-wireless systems to function without a clear line of sight between the transmitter and the receiver.
Dropping the ball
The fixed-wireless broadband access market is segmented by spectrum allocation, with services operating in a number of bands. The highest-profile activity has taken place in the 2.5-GHz MMDS (multichannel multipoint distribution service) band. Fueled by the FCC's 1999 decision to open up the MMDS band, which was originally slated for wireless delivery of cable TV, to two-way communications, Sprint and Worldcom threw their hats into the ring. Also with the aim of entering the local loop with voice and data services, AT&T Wireless, under the moniker Project Angel, invested in fixed-wireless technology and services. At the height of the optimism, big names in the convergence arena, including Cisco Systems, also placed bets on the technology.
With three major service providers signed on, one would have expected fixed wireless to take off in a big way. Today, however, the landscape looks bleak. AT&T recently sold its Project Angel assets to broadband equipment provided Nexio, and Cisco has pulled out of the market. Meanwhile, both Sprint, which targeted the residential market, and Worldcom, which set its sights on the business arena, have ceased their rollouts of fixed-wireless services.
“The first-generation equipment...just didn’t meet the expectations of the market from a user-experience, installation, and cost-model perspective. It just wasn’t mature, and so we are suffering from this black eye.”
Andy Kelm, Nokia
At a recent meeting of the Wireless Communications Association (WCA), Sprint identified cost and installation as the key challenges it faced and claimed that available products were not up to the task of delivering coverage in all its markets, according to Michael Greenbaum, president and CEO of Hybrid Networks, which supplied its fixed-wireless system to Sprint's venture.
"Line-of-sight technology was found wanting in the urban markets that dominated Sprint's footprint," Greenbaum says. The technology worked well in wide-open flat areas without foliage, but Sprint ran into problems in areas with more hills, according to Greenbaum. In addition, pine and fir trees, whose needles are approximately the same length as the MMDS signal wavelength, caused problems. Sprint subsequently challenged the industry to come up with a system that could deliver above 85 to 90 percent coverage and be self-installed by the consumer for under $500. "That's where the market is right now," Greenbaum says.
"Sprint said it would deploy in 50 markets by the end of 2000, and ended up stopping after 13, while Worldcom said it would go into 40 markets and after 14, it stopped," says Asif Naseem, vice president of business development at Iospan Wireless, a provider of wireless communications products and services based on multiple-antenna technology. "In those markets, they also gained experience as to what the potential issues were as they added subscribers. The three major issues were cost, capacity, and coverage."
Focused on the consumer market, Sprint found that it had to do three truck rolls on average to qualify a single subscriber, at a cost of anywhere from $1000 to $1600, according to Naseem. "Every time it rained and the channel characteristics changed, or something went wrong with the CPE [customer premise equipment], such as the wind or a child changed its direction, this required a truck roll," he says. "After a few years' experience, Sprint found that the truck rolls were breaking their business case. It was taking them 2.5 to three years to just break even and recoup the cost of acquiring the subscriber."
Moreover, the user experience wasn't all that great, with problems reminiscent of those associated with DSL installation. Customers not only had to wait for weeks for the truck roll, but also—in as many as one out of three cases—found that they could not get service, Naseem says. "The current-generation equipment had direct line-of-sight requirements between transmit and receive," he says. "And with that, the denser the area, the less coverage. Critical data shows that in an average neighborhood, Sprint was getting no more than 30 to 50 percent coverage. That provided a marketing challenge for a service provider. And so, non-line-of-sight technology becomes a primary requirement."
The third issue plaguing first-generation technology was capacity limitations, Naseem says. "If you use all your available spectrum in a market, you are capacity limited, he says. "And if you run out of capacity, you can not reuse the spectrum unless you have interference algorithms for spectrum reuse."
Iospan aims to address these issues with its AirBurst technology, which utilizes multiple-input, multiple-output orthogonal frequency division multiplexing (MIMO-OFDM) to improve spectral efficiency. Based on cellular architecture and smart-antenna technology, AirBurst provides extended coverage, increases capacity, and provides quality-of-service features. The system helps service providers reduce the amount of infrastructure deployed while offering greater coverage than current systems, Naseem contends.
The company also claims that AirBurst offers residential and business customers multimegabit access via user-installable, desktop CPE. "For truck rolls to stop, CPE has to be user installable," Naseem says. "And that means it has to be an indoor install, not on the top of the roof or outside the house, which begs a professional. Since the subscriber unit has to be inside the house, it has to work in a non-line-of-sight environment. We have developed a system that addresses those requirements."
Iospan offers AirBurst chipsets, reference designs, and board-level systems. "We will sell to broadband system vendors who will then take our technology and commercialize products," Naseem says. "We have implemented reference designs at both ends of the network." The company has also begun a field demonstration of its technology in the San Francisco area using the MMDS frequency band. "We are working with service-provider technical teams and with system vendors, and we are bringing the two together," Naseem says.
Band aid
At BeamReach Networks, another fixed-wireless startup, vice president of marketing Bernard Aboussouan agrees that first-generation technology did not make sense from an economic point of view. "If you look at Sprint and what they are trying to offer, it's basically an alternative to DSL or cable modem," he says. "Yes, they were able to burst at a higher bit rate, but if you're charging the same tariff rate as some of the other large competitors, you need to have a technology that's ultimately less expensive. By the time they counted the capital cost of the equipment, the multiple truck rolls, and the installation costs, there was no way they could make money."
Aboussouan indicates that a number of elements make up a viable business proposition, and non-line-of-sight capabilities and self-installation are key. Because line-of-sight covers only a certain percentage of the users in an area, a certain percentage is automatically excluded, he points out. "You have to be able to cover more than 90 percent of the subscribers—and reliably, with good availability, where the link is connected a very high percentage of the time," he says. "And then, you have to address installation, which has to be simple enough that the customer can install it and the operator doesn't need numerous truck rolls."
“For truck rolls to stop, CPE has to be user installable. And that means it has to be an indoor install, not on the top of the roof or outside the house.”
Asif Naseem,
Iospan Wireless
BeamReach aims to solve these problems with adaptive beam-forming technology, which, the company claims, delivers both wide coverage and high spectral efficiency in supercell and multicell environments. "We are able to form a beam on a per-subscriber, per-packet basis in an adaptive way," Aboussouan explains. "We can also form multiple beams at the same time and address multiple subscribers within the cell. Moreover, we form that beam very directionally, toward each subscriber. That extends the coverage, reduces interference, and provides high frequency reuse. Because we can use the frequency over and over from a given cell and a limited amount of spectrum, we can serve a lot of subscribers."
While targeting all frequencies below 6 GHz, including MMDS, BeamReach is demonstrating its system in the WCS (wireless communication services) band at 2.3 GHz. Since this is considered one of the most challenging bands, it allows the company to prove the viability of the technology in every band. "Because we provide spectral efficiency and are able to reduce interference, we can operate in that narrow band and make it valuable," Aboussouan says. The band's license holders include some major telcos such as Verizon and BellSouth, as well as Worldcom, and AT&T Wireless.
Out of band
While eliminating line-of-sight requirements and enabling self-installation will certainly help improve the fixed-wireless business case, service providers will need to roll out other services, such as voice, to make the venture pay, according to Yatish Pathak, president and CEO of SOMA Networks. "With all of the money made off of voice services today, no one was offering carrier-quality voice," he says. "Trying to make a data system without a robust, high-quality voice offering will kill a carrier. It becomes a commodity war at $20 to $30 a month. It has to be broadband data so you can compete with DSL, has to be carrier-class voice, and has to be a converged network."
In this vein, SOMA has licensed Qualcomm's CDMA (code division multiple access) technology to deliver the SOMA Air Interface. "Earlier-generation fixed-wireless systems for data rely upon line-of-sight engineering with expensive customer equipment and costly installation, while offering limited network capacity," Pathak says. "Our next-generation fixed-wireless system leverages CDMA strengths in cellular deployment, network scalability, spectral efficiency, and trivial frequency planning with spectrum reuse between sectors and cells."
“You have to be able to cover more than 90 percent of the subscribers—and reliably, with good availability, where the link is connected a very high percentage of the time.”
Bernard Aboussouan,
BeamReach Networks
The SOMA Air Interface builds on CDMA, but is optimized for fixed multimedia broadband applications, with higher peak data rates, aggregate capacity, and spectral efficiency. SOMA Networks says that its Amosphere architecture combines wireless technology, QoS-enabled IP, and distributed computing, while its CPE hardware, SOMAport, can be installed and provisioned by end users. The Amosphere software-defined service environment enables multiple carrier-grade voice lines, advanced calling features, and high-speed Internet access.
"People get hung up on whether it's CDMA, OFDM, or even FM," Pathak says. "It really doesn't make a difference. It is only a tool. We settled on the 3G radio interface as the physical layer and then built on that, solving the issues of channel capacity. All of a sudden, the business case makes sense in terms of the kind of revenue you can drive from each user, because you can get voice and data and all kinds of other services."
Other opportunities
With the big players like Worldcom and Sprint finding limitations with first-generation technology, it's not surprising that suppliers of first-generation products are today crying "sour grapes" in regards to the potential mass market. These companies, including the likes of Hybrid Networks and Vyyo, however, have far from bowed out of the fixed-wireless arena.
"Sprint and Worldcom made large investments in fixed-wireless broadband, and were always trying to answer the question, 'Is the technology economical and is it scaleable to the level that is needed to serve large numbers of people on a very reliable basis?'" says Hybrid's Greenbaum. "We can say that the technology of our present-generation systems answered those questions with flying colors."
In fact, vendors like Hybrid have found success with smaller US service providers and in the international market. "The real interesting thing to me is that this is not a technology that is uniquely for the US market," Greenbaum says. "The international market presents an opportunity far greater than we have in the US. The reason is that low-cost, high-quality infrastructure or wireline platforms are not available if you go to Asia, Africa, or Latin America. We have quite a number of customers who are going to continue to invest this technology, and we will be a first-rate supplier."
Existing suppliers also caution that the claims of second-generation upstarts should be taken with a grain of salt. Second-generation systems are just beginning to make their debut, and some are farther along than others in the testing and trial process. Greenbaum questions whether such systems will be able to pass muster in the long run. "Some of the hype that some operators have placed on self-install is a little bit premature," he says. "No one that I know is able to deliver that without excessive cost. It's a very complex system, and all of the parts have to work together very well to deliver the kind of reliability we have already delivered."
“The international market presents an opportunity far greater than we have in the US. The reason is that low-cost, high-quality infrastructure or wireline platforms are not available if you go to Asia, Africa, or Latin America.”
Michael Greenbaum,
Hybrid Networks
Vyyo senior vice president of strategic relations Arnon Kohavi puts the point more strongly: While technical challenges such as non-line-of-sight and self-installation are important to grow the market, they're not the whole story. "We are developing non-line-of-sight, but we don't believe that by itself will solve the problem," Kohavi says. "Self-install is complex and expensive, and it requires that houses be very close to each other. It needs up-front capital spending for carriers to deploy it. With any new technology, there is an evolutionary process, and the cost goes down as volume goes up. But no one will get the cost down when there is no volume."
Consequently, Kohavi asserts, fixed-wireless broadband access has lost its chance to become a mass-market residential service in the US. "While a year ago, Sprint, Worldcom, and AT&T were very strong promoters for the new technology, today, there is no catalyst to a make this a mass-market deployment," he says. "There is a big international opportunity where there is no infrastructure of DSL and modems. In niche markets, it still can be done, but these dreams of millions of subscribers in US cities will not happen."
Mass-market maybe
Others haven't given up on the dream. Andy Kelm, head of marketing in the Americas for the Nokia Wireless Routing group, agrees that the failures of AT&T, Sprint, and Worldcom didn't bode well for the industry. "Nobody saw the massive deployments they expected," he says. "The first-generation equipment—line-of-sight, point-to-multipoint equipment—just didn't meet the expectations of the market from a user-experience, installation, and cost-model perspective. It just wasn't mature, and so we are suffering from this black eye."
However, Kelm is confident that fixed wireless is a viable play in the broadband market. "The promise of broadband is still as large as it ever has been," he says. "Subscriber numbers are not slowing down, and demand is still robust. At the same time, carriers are backing off on their wired deployments because of capital budget restraints. So we're seeing renewed interest from WISPs [wireless ISPs] and carriers. The next generation of non-line-of-sight products, of which ours is a part, does meet customer needs in terms of services provided and business model. There was just a lot of hype that the industry wasn't able to live up to. But it's now on track to do so."
Nokia offers the 2.4-GHz RoofTop Wireless Routing system, a wireless mesh network that, according to the company, solves the line-of-sight difficulties typically associated with residential neighborhoods. Each node requires only a connection to neighboring nodes, not all the way to a central basestation. Each subscriber is a full IP router, routing traffic around trees, hills, and buildings and enabling deployment in diverse locations, according to Nokia.
Despite his optimism, Kelm acknowledges that without the big players on board, the fixed-wireless arena will be relegated to niche status for some time to come. "This market will not happen without large carriers, but there is an evolution path," he says. "It can survive for a while on CLECs [competitive local exchange carriers] and WISPs, while the carriers are a little slow to spend. There are a lot of startups promising all kinds of technology, and so the market has to digest those technologies and see which ones are viable and which aren't. There is still a lot of noise out there, so carriers are sorting through what's real and what's not."
BeamReach's Aboussouan agrees that the major carriers will not jump back on the bandwagon overnight. "There is a lot of suspicion in the market right now," he says. "If you look at all of the failures, the market has to get confident that the technology can be deployed economically and does make sense." A marginal improvement won't do the trick, he continues; only a significant improvement that makes it clear to operators that they can make money deploying fixed wireless.
AI
