Consumer Friendly
New MMDS Technologies Hold Promise of
Consumer-Priced Broadband Wireless
By Charlotte Wolter
Posted: 02/2000
Although much of the action in broadband fixed wireless has been on high-frequency, high-bandwidth technologies, such as those in the local multipoint distribution service bands, another segment of the industry, below 5GHz and aimed at a consumer audience, is heating up in the first few months of 2000.
Two new entrants in the technological race to find winning fixed broadband
wireless solutions have focused attention on the consumer and small-office
space. Cisco Systems Inc. (www.cisco.com) has
developed an innovative approach to MMDS transmission called vector orthogonal
frequency-division multiplexing (VOFDM). AT&T Corp. (www.att.com)
announced it finally will deploy the 1.9GHz fixed wireless technology it has
field tested for several years.
However, so far broadband wireless has been more promise than payoff, as
deployments have been limited by the slow development of products and the lack
of maturity in the technology. Individual technologies do not necessarily
provide overall solutions to
service providers who must deliver a range of services and bandwidths to
customers.
Some have predicted 2000 will be the year of broadband wireless, but doubters are numerous, some suggesting that rather than escorting in the new millennium, broadband wireless is more likely to wait one or two more years to take its place on stage.
AT&T Chairman and CEO C. Michael Armstrong had the financial analyst community abuzz as he announced in briefings early in December that AT&T would finally deploy the fixed broadband wireless technology it originally announced in February 1997. The service, called Project Angel, will roll out in 2000 in three markets: Dallas; Ft. Worth; and an unnamed third location. A market trial with 100 paying customers is already under way in the Dallas area.
Project Angel will spin off from the parent AT&T along with the mobile wireless services in the initial public offering (IPO) of a wireless tracking stock the company announced late in 1999.
AT&T Takes to the Airwaves
Project Angel, using a proprietary technology AT&T developed in-house, transmits in the 1.9GHz band, where AT&T has nationwide licenses for just 5MHz to 10MHz. The company plans to deliver up to 384kbps of always-on Internet access (with choice of ISP) and two phone lines to residential and small-business customers. Cell sites will have a radius of 50 kilometers or about 35 miles.
AT&T has been highly circumspect about the technology itself as well as its deployment plans, but a few details are available. "The way they're leaking it out is Michael Armstrong is talking to financial analysts," says an industry analyst speaking off the record. "He talked it up in general terms, and AT&T never really made a formal declaration of what's going on with it."
"Clearly AT&T wants to come in and take over the local loop, and they are trying to reach out with new technologies, cable and wireless," says Bob Egan, vice president, mobile and wireless group, The Gartner Group Inc. (www.gartnerweb.com).
AT&T reportedly found the cost of its initial technology too high in early testing, about $750 per household, which is far above the $300 to $400 for a cable upgrade to data (telephony is about $100 to $150 more). After a retooling, the company apparently was confident the cost trajectory was acceptable, which made it willing to move toward deployment. One factor in favor of wireless is that many expenses are delayed until a customer actually comes on a network, unlike wired, where the entire infrastructure must be built up front.
"It appears they were testing a couple of things," says Jeanette Noyes, research manager, International Data Corp. (www.idc.com). "One is using the fixed wireless network for a mix of voice and data services. In some [other] scenarios, there could be a convergent mobile and fixed application where a customer would be mobile when they went out and at home on the fixed network. The advantage is to take the load off the [mobile] wireless network. It is unclear whether the underlying strategy is fixed wireless or if that is just one variation."
Noyes found the cities where AT&T has suggested it will deploy Project Angel are also cities where AT&T has a strong cable presence. The company's strategy likely will vary considerably depending on what other resources, cable and wireless, it has in a metropolitan area. "In some cases the cable network will support telephony and in some it will not support it right away. So they can do a kind of triage," choosing between cable and wireless to provide service. "Ultimately what they want to do is offer a bundle of voice, video and data to customers," Noyes says.
The company is working with Ericsson Inc. (www.ericsson.com) to develop the technology, says Egan. A company spokesman says AT&T plans to add other third-party providers in the future.
Egan believes there are significant barriers to AT&T's rollout, not the least of which are bandwidth shortages in its existing wireless services. AT&T has been seeking spectrum relief at 1.9GHz (for its mobile service now at 800MHz). The company has been buying infrastructure from Ericsson that allows it to migrate mobile customers from 800MHz to 1.9GHz when it runs out of capacity, Egan says.
"So you have to ask yourself, how many of those markets won't be served by fixed broadband because there's not lots of spectrum. It's just 5MHz to 10MHz in those markets and they have to use some to support existing voice users," Egan says. "The bottom line is that when it comes to broadband wireless, I am very skeptical about any data solution from AT&T," despite the company's large investments. Not only is the company strapped for bandwidth but it will face competition from Sprint Corp.'s (www.sprint.com) and MCI WorldCom Inc.'s (www.wcom.com) developing MMDS-based services, both of which have much more spectrum.
Egan feels there still is a chance for success and says that could depend on AT&T looking to technology that is more bandwidth-efficient than its current time-division multiple access (TDMA)-based scheme. The company has eschewed systems based on QUALCOMM Inc.'s (www.qualcomm.com) code-division multiple access (CDMA) technology, but it is now working with Ericsson, which owns QUALCOMM's infrastructure business. QUALCOMM has a CDMA technology called High Data Rate (HDR) that can deliver up to 1.8mbps. "It's pretty likely that AT&T for a fixed broadband solution may have to use a different technology, one that is not aligned with a mobile network," Egan says.
VOFDM in the Spotlight
When Sprint and MCI went on a buying spree of MMDS spectrum in 2000, they revitalized a frequency band that had been idled by lack of development funds and withering cable competition. They also sparked a rebirth of technological development as companies, such as Hybrid Networks Inc. (www.hybrid.com) and Spike Technologies Inc.(www.spiketechnologies.com), found new customers or infusions of cash. Sprint invested more than $11 million in Hybrid Networks and placed an additional $10 million order for equipment.
MMDS has price advantages that make it suitable as a consumer service technology. It has a range of up to 50 kilometers, and low-cost radio frequency (RF) equipment is available, especially for CPE, because of the technology's long history as a one-way video transmission medium.
In late October, Cisco entered the MMDS picture with its innovative VOFDM modulation technology that could make transmissions in this MMDS spectrum (2.5GHz in the United States) more robust and resistant to interference. The proposal was backed by 10 other firms, including Motorola Inc. (www.mot.com), Cisco's partner in an LMDS broadband wireless technology venture called SpectraPoint. Other partners, Broadcom Inc. (www.broadcom.com) and Texas Instruments Inc. (www.ti.com), will make key application-specific integrated circuits (ASICs) and digital signal processor (DSP) chips for the system.
The VOFDM proposal is significant because it is the first attempt to standardize an overall MMDS transmission system, although MMDS vendors were moving toward employing the cable industry's data over cable service interface specification (DOCSIS) standard for CPE.
Cisco's proposal to use VOFDM is an attempt to deal more effectively with one of the most difficult technical problems in using MMDS, one that has limited its deployment. The advantage of MMDS is its long 50-kilometer reach. The disadvantage is its susceptibility to fading and interference where there is a lot of foliage, or worse, wet foliage. It also has had stringent line-of-site requirements.
VOFDM modulation is a variant of the technology used to transmit digital terrestrial TV in Europe. It has several capabilities that make it potentially superior to the quadrature amplitude modulation (QAM) now widely in use in MMDS systems. It is resistant to multipath interference, which is caused when signals bounce off large objects, such as tall buildings or mountains, and arrive at a receiving antenna at slightly different times, sometimes canceling each other out.
"At MMDS frequencies, two things are working against you in multipath," says Allan Evans, director of product management, Netro Corp. (www.netro-corp.com), a vendor of LMDS equipment. "One is the longer distances involved, and, two, wider antenna beams. A subscriber antenna at 2GHz may have a 20-degree beamwidth where LMDS has just a three-degree beamwidth. So there is less opportunity at higher frequencies for reflected signals to come into the main beam or interfere with beam path."
There have been attempts to use an adaptive equalizer to deal with the problem, but as the time delay in multipathing gets worse, the adaptive equalizer must be more complex, and more expensive. VOFDM gets around that by modulating the signal onto as many as hundreds of smaller carriers in the signal rather than one big one, and the receiver recombines the carriers into one.
"If you had VOFDM on a wire, you would call it DSL," says Wesley Vivian, a well-known consultant in wireless technologies. "They are the same animal. In VOFDM, you break up the signal into a lot of little parts, each one of which has lower modulation on it, and send all the pieces separately and put them back together at the other end, and that is what DSL does."
Vivian says the VOFDM approach has potential to be successful with multipath, "though an equalizer can do it almost as well." As for dealing with trees, a problem that becomes more acute as leaves wave in the wind and scatter the signal, "with a good VOFDM signal, you can get rid of a lot of the waving in the wind problems, but it is hard to get rid of the fact that the signal can get so weak that you can't get it."
In addition, at this point, VOFDM technology can be more expensive than QAM, although Cisco's cooperation with Broadcom is intended to reduce costs by creating VOFDM ASICs. Also, the inclusion of Motorola, which makes DOCSIS modems, is a further effort to make its entire system less costly. "It remains to be seen whether VOFDM passes the 'so-what' test," says Evans, meaning, whether the improvements in performance make it worthwhile to switch to a new technology.
Cisco has proposed to license the new technology free. At press time though, none of the existing MMDS manufacturers, such as Hybrid Networks or Spike Technologies, had announced intention to make equipment based on it. However, these companies are very likely to step up to the plate on the new Cisco system, Egan says, "especially a lot of the DOCSIS companies. These companies are diving in with that standard and Cisco has investments in that area. Those making DOCSIS modems could make [consumer equipment] for this, including Hybrid or Spike."
Sprint, which also has an investment in Hybrid Networks and sits on its board, is evaluating the technology at its labs in Burlingame, Calif. A contract from one of the two new big-time players in MMDS, Sprint or MCI, would be key to winning broad industry support for the proposal.
"Sprint has a two-phase rollout plan," says Egan. "The first is around the Hybrid Networks solution, leveraging whatever modem vendors happen to be available. Phase two is for higher speed services and much larger money to be awarded to any equipment supplier, and that is where Sprint is looking at Cisco among a short list [of] candidates, particularly with VOFDM."
MMDS deployments have not been limited to residential, though it remains the focus of applications of the technology. Using traditional QAM-based MMDS, Spike Technologies was awarded a contract on Dec. 20 by Oxford Telecom Inc. (www.oxfordtelecom.com) to deploy its HighPoint 2.5GHz MMDS system in Portland, Maine, in the first quarter of 2000.
Oxford will use this system to provide high-speed, two-way services to businesses and consumers. Services include Internet access, VPNs, data vaulting and e-commerce applications.
