ArrayComm Deal Boosts Fixed-Wireless

The role of fixed wireless as a challenger to cable and digital-subscriber line in the residential high-speed-data market was further solidified last week with a decision by Sony Corp. of America to back the ambitious national-network plans of ArrayComm Inc.

Sony's $8 million contribution to ArrayComm's "i-BURST" development project was just a small piece of the $2 billion or so ArrayComm and its partners will ultimately need to spend if they are to meet deployment goals of reaching 60 percent of the nation's households.

But the commitment from the consumer-electronics and media giant marks a new emphasis on fixed wireless as a means of driving its new service ideas into the marketplace, said Yang Hun Lee, executive vice president for corporate strategy at Sony.

"We now see a growing importance of a new networked environment for the distribution of consumer applications at extremely low cost using the Internet," Lee said. "Sony's investment in ArrayComm's i-BURST technology represents an opportunity for us to propel the development of a technology that will enable the delivery of new, media-rich wireless services built around our own core content and technologies."

Sony's plans for use of high-speed wireless data "impact on every division of the company," said a source close to the company who asked not to be named. "The emphasis at all levels is on network connectivity, and they view fixed wireless as a very low-cost means of delivering games, music, video and other content they supply to the consumer appliances that come out of their manufacturing divisions."

ArrayComm is telling investors it expects to supply 1-megabit-per-second data access to end-users at a monthly cost comparable to what consumers pay today for dial-up access. "Once we get to high volumes of users, the monthly fee will be about $22," ArrayComm CEO Martin Cooper said.

The Sony backing-part of a new $15 million round of financing for i-BURST-is the first of many developments in the push to get the new platform into operation, Cooper noted. "We're very close to naming a carrier for the project, and there will be other partners, as well," he said.

ArrayComm expects to soon win an experimental license from the Federal Communications Commission to launch a market test involving "up to 3,000 paying customers" in an unnamed city by mid-2001, Cooper said. "We'd like to obtain 10 megahertz of spectrum, although we could work with 5 MHz if we have to," he added.

With 10 MHz, the ArrayComm system will be able to pump out data to a cluster of end-users at 40 mbps. Each user within reach of the neighborhood cell site would receive services at 1 mbps or better, Cooper said.

The system uses adaptive smart-antenna technology based on spatial-processing software to achieve high bit rates over relatively small segments of spectrum.

"Our IntelliCell'technology is a hardened, proven technology now deployed in networks serving over 3.5 million customers worldwide, but there are other technical areas of development we're working on for the i-BURST system," Cooper said.

Whereas the current deployments are for voice applications in various types of wireless systems, the new application is designed for an all-IP (Internet protocol) distribution strategy that requires special adaptations for RF propagation, he noted.

"We're partnering with Redback [Networks Inc.] to develop the means of enabling high-speed wireless Internet access," Cooper said. "This is very different from the techniques we use with fixed-wireless local-loop voice applications."

ArrayComm hopes the FCC will ultimately license 10 MHz of spectrum at or below the 3.5-gigahertz level-"preferably somewhere in the 2-GHz range"-for the i-BURST system, Cooper said.

The plan calls for a single carrier to take responsibility for building the national network, but the technology could also be used by other spectrum holders to build competing networks, he noted.

Sony's backing of ArrayComm comes in the wake of AT & T Corp.'s commercial launch of its fixed-wireless IP-voice and data service in Fort Worth, Texas, to be followed by rollouts in two other cities later this year and many more in 2001.

AT & T's system is designed to deliver four lines of voice and data on a shared-access basis at 512 kilobits per second, with plans calling for gradual expansion of the data rates into the multimegabit range over time.

Rather than using advanced antenna technology, like Array-Comm is doing, AT & T's system relies on an advanced modulation system known as "orthogonal frequency-division multiplexing" and sectorization of its antennas into four quadrants at each cell site to achieve high levels of spectral efficiency.

"We allocate a total of 8 MHz out of the available 10 MHz to data, and break that into 2 MHz per sector, 1 MHz for each direction," AT & T senior vice president Larry Seifert said.

This leaves plenty of spectrum available for raising the bit rate. "We plan to go to 1 mbps per sector in each direction by midyear," he noted.

The AT & T system is designed to accommodate high voice-usage spikes in real time by tapping into the spectrum available for nonvoice data, Seifert added. "Although we're dividing the spectrum between voice and data, we can dynamically move channels when necessary," he said.

While AT & T's $750-per-customer cost is low enough to meet its pricing needs at this early stage of deployment, the company and its competitors on the wireless side are headed toward much lower costs, as seen in a test of another technology platform now under way in Chicago.

There, Lucent Technologies is working with IP-technology developer Komondo Inc. to fine-tune a fixed-wireless platform that an unnamed carrier plans to put into operation for a marketing trial this summer, according to test participants.

"Lucent has told us that the customer-premises equipment [excluding antenna] has to be priced at under $100, and that's what we've accomplished," said Steve Toteda, vice president of Komondo, a Silicon Valley start-up.

The Lucent system is designed to work in all-IP mode for voice and data, which is what companies like MCI WorldCom Inc. and Sprint Corp. want in efforts to use multichannel multipoint distribution service and other spectrum categories to deliver local services to the mass market.

While the trial planned this summer will operate in the MMDS tier at 2.5 MHz-close to the personal-communications-services tier at 1.9 MHz-the architecture to be used for the MMDS application is similar to the one used by AT & T, Toteda said.

Small base stations will be positioned to serve a few thousand households, reducing the number of users contending for bandwidth to the point where each can receive a full complement of broadband services, including video, he noted.

"AT & T has done what it said it would, but it's just the beginning of what you'll be seeing as the year goes on," Toteda said.