Fixed-position wireless: broadband without the wait

Fixed-position wireless:
broadband without the wait

CommVerge
by Gil Bassak, Contributing Editor
June 1, 2000

By taking a wireless route around the last-mile bottleneck, two emerging services may finally end what for many is the great wait for high-speed network access.

One option, the 2.5-GHz multichannel multipoint distribution service (MMDS), is seen as a wireless alternative to DSL and will offer bandwidth of 2 Mbits/sec or more. The second option, the 28-GHz local multipoint distribution service (LMDS), promises to deliver much higher throughput, starting at about the T1 data rate of 1.54 Mbits/sec and extending to 155 Mbits/sec or more for businesses that want the throughput of, but don't have access to, optical fiber.

Unlike forthcoming wideband cellular services, which target mobile communications, MMDS and LMDS take aim at customers in fixed locations, who today depend on the reach of copper or fiber cables. And the market seems to be there. One recent study by Allied Business Intelligence says that MMDS and LMDS will gain 9 million subscribers by 2005, with MMDS getting 70% of the residential and small office/home office market, while LMDS evolves into a major wireless broadband technology for business. Other estimates are much higher.

The appeal of these wireless services comes as much from the fast pace of setting up a network and providing service as from their high data rates. In contrast to the often slow and troublesome deployment of wired broadband services, wireless installations are easy.

“Delivery of service is a simple matter of putting an antenna on the roof and connecting it.” Ray W Nettleton, Formus Communications

"As rooftop to rooftop networks, we can build them very quickly," says Ray W Nettleton, a senior vice president and chief technical officer for Formus Communications. "Moreover, delivery of service is a simple matter of putting an antenna on the roof and connecting it." Formus is unrolling LMDS across Europe, where it offers Internet access and virtual private networks with data rates up to 8 Mbits/sec.

For service providers, wireless networks mean much lower costs. "If I have to dig out a neighborhood to put in a cable infrastructure, that's a lot of capital expense before the customer gets his first Internet message out," says John Ehler, a marketing manager for Cisco Systems, which supplies both MMDS and LMDS systems. Indeed, Peter Rysavy, who heads Rysavy Research, estimates that in an urban environment, running just one mile of optical fiber costs as much as $250,000. No doubt that explains why only a small percentage of buildings have fiber, leaving the door wide open for a low-cost alternative.

Still, Rysavy sees MMDS and LMDS more as alternatives for places where wire and cable are simply not available, rather than as direct competitors to traditional wires. Right now, "cost per subscriber is still too high to compete head-to-head with wire-line technology," he notes. However, by 2005 Rysavy guesses that the two wireless services combined might garner 25% of the fixed, or terrestrial, broadband market.

To be sure, the die is not cast. MMDS and LMDS face competition from other broadband wireless services. These alternatives include the 38-GHz band, which is licensed for general-purpose communications services, portions of a relatively newly available band at 5.7 GHz called the Unlicensed National Information Infrastructure (UNII), and the unlicensed slots of the ISM (industrial, scientific, and medical) bands at 2.4 and 5.8 GHz.

Still, for technical and practical reasons, both MMDS and LMDS have generated considerable interest among equipment providers and carriers ever since the FCC authorized them for two-way digital communications and put them on the auction block.

Humble beginnings

MMDS started life years ago as a one-way channel for transmitting cable television to customers not served by a copper connection. Seen then as an "industry stepchild," according to Carlton O'Neil, vice president of sales and marketing for system builder Ensemble Communications, MMDS "sent many companies into bankruptcy multiple times." But in September 1998, the FCC opened it for two-way voice and data communications, attracting industry attention.

Because of their relatively low frequency, MMDS transmissions cover an impressive radius of about 50 miles in theory, 35 miles in practice. Although MMDS is considered a line-of-sight technology, it differs from LMDS and other millimeter-wave services, which have about a 2-mile radius, in that its signal bounces off of obstructions rather than being absorbed by them.

Normally, these multiple signal reflections bouncing off of buildings and foliage—called multipath—pose a problem on the receiving end. But one equipment vendor, Cisco Systems, has found a way to combine multiple reflections into one strong, coherent signal using a technique called vectored orthogonal frequency division multiplexing, or VOFDM. However, others say this technique unnecessarily wastes bandwidth and that multipath can be handled more simply while preserving bandwidth (see sidebar, "Scheme schism").

Regardless, MMDS' large radius means that, in theory at least, one transmit and receive station is all a service provider would need to cover an area with a 35-mile radius. In reality, however, to achieve the high capacity needed to serve dense markets, several cellular sites would have to be installed within that area, allowing frequencies to be reused. Still, notes Carroll McHenry, chairman and CEO of service provider NuCentrix Broadband Networks, "covering an area with one to five cell sites is still a lot more economical that stringing coax, fiber, or copper across those kinds of distances."

NuCentrix is running one-way MMDS trials in Austin and Sherman, TX, and expects to have two-way authorization from the FCC in the fourth quarter. The company's initial offering will be Internet access, which could be available as early as the fourth quarter. However, McHenry is not ruling out subsequent rollouts of Internet telephony and high-quality video. He also sees a place for MMDS beyond fixed-location access, saying it can serve portable devices until cellular and PCS systems are able to deliver on their own promise for broadband service.

"Over the next two to four years, cellular and PCS will have to struggle mightily just to get to 256 kbits/sec," McHenry says. "The MMDS spectrum has the potential to offer true broadband capability—about 1 to 1.5 Mbits/sec—to portable computers." He cautions that because MMDS lacks a cell handoff mechanism and requires a notebook-sized antenna, this approach is unsuited for true mobile applications. But, he adds, there's no reason why it wouldn't work if you were sitting at a desk or in an airport.

As it is, NuCentrix is a relatively small player in the MMDS market, which is dominated by MCI Worldcom and Sprint, two giants that are set to merge. Last year, for instance, MCI Worldcom spent $4 billion to acquire four MMDS providers. Similarly, Sprint has been buying up broadband wireless companies across the nation; so far, the firm has acquired six, and with them formed its Broadband Wireless Group.

Combined, the two companies say their MMDS services could reach 60 million US homes in 100 markets. Customers would be able to buy broadband Internet access and, eventually, telephone service that circumvents the Regional Bell Operating Companies. "They see this a huge opportunity to bypass the local telephone companies," Rysavy says. "So I think you will see fairly aggressive deployments by 2001, though it won't represent a significant chunk of the market until 2002 and 2003."

Both companies are completing trials in some cities and starting others. For example, during March and April, MCI announced that it was conducting MMDS trials in Jackson, MS, Baton Rouge, and Memphis, with plans to begin tests in Dallas/Ft. Worth and Boston. In those trials already in progress, says spokesman Joe Paluska, the technology is capable of bursting up to 10 Mbits/sec, although service is limited to "the sweet spot" of between 128 kbits/sec and 2 Mbits/sec in both directions. The service costs up to $600/month, Paluska says, depending on the bandwidth.

In Jackson, MCI is also testing a 310-kbit/sec wireless LAN service for the 300 residents of an apartment complex. There, tenants pay a refundable $100 deposit and get an add-in card for their PC. In addition, there's a $39.95 monthly fee for unlimited use. In this case, MMDS delivers the data stream to the building and a wireless LAN completes the link to the customer's PC.

Although MCI is presently focusing on MMDS to deliver data, future plans call for adding voice. Admittedly, Paluska says, "the technology for delivering voice is not here yet." But he expects that to change in the next 12 to 18 months. MCI is already planning limited testing of telephone service in Dallas. "Anytime we can bypass the local exchange carrier, it saves our company, and therefore our customer, money," Paluska says.

For its part, Sprint has gone beyond trials. As this article went to press, the company was set to roll out its first MMDS service, Sprint Broadband Direct, in May in the Phoenix area. "It's a two-way, high-speed Internet product," says spokesman Robert Hoskins. Available to both businesses and consumers, the service will provide a downstream burst rate of about 5 Mbits/sec, though that could drop to as low as 2 Mbits/sec when the system is fully loaded. Upstream data will travel at 256 kbits/sec.

Consistent with the promise of wireless broadband, Sprint expects that, once it starts offering the service in an area, customers will wait only five to seven days between placing an order and getting online. By the end of the year, Sprint hopes to bring Broadband Direct to 20 markets nationwide.

Strictly business

Where MMDS looks to the home and small businesses for its customers, LMDS focuses strictly on the high data-rate demands of business. "LMDS is for those small to medium size businesses that need more bandwidth than a T1 line, but not as much as fiber," says Kimberly Tassin, director of marketing for Wavtrace, an LMDS equipment maker. Others also see multidwelling residences as likely LMDS candidates.

Moreover, although companies are looking to LMDS for high-speed Internet access, they also want it, as is typical with fiber customers, for their telecommunications. "With LMDS, because the large corporate customer would just as soon get a broad pipe for all its services, you will see more or less an even split between Internet and telephony services," Rysavy says.

“If I have to dig out a neighborhood to put in a cable infrastructure, that’s a lot of capital expense before the customer gets his first Internet message out.” John Ehler, Cisco Systems

As mentioned, LMDS is one of several millimeter-wavelength bands available for fixed broadband access and the most recent to be made available by the FCC. As LMDS trials begin, the big issue for some seems to rest on the ability of equipment manufacturers to produce high-performance, low-cost systems. "There's been a lot of innovation and experimentation going on, but there's a big difference between demonstrating something and actually being able to deliver it," Rysavy says.

Indeed, early systems suffered from limitations like fixed frequency allocations and low data rates, according to Ensemble Communications' O'Neil. "The dirty little secret of the LMDS industry is that there are a lot of slides that said what should happen, but the systems didn't really do those things," O'Neil says. "They were a kind of reinvention of point-to-point radios, which have been around for 50 years." Consequently, carriers were reluctant to go beyond the trial phase. More recently, O'Neil notes, improvements in data rates, bursting ability, and cost have brought the systems more in line with market needs.

Another consideration for LMDS is that, given its 2-mile, strictly-line-of-sight range, it is justified only in densely populated urban areas. That should not be a problem, however, because the tallest buildings are likely to be the ones where you will find an optical fiber. Given their relative height, these would be the buildings best suited to house the base station and antenna needed to give access to subscribers.

With their promise of hopping over the local-loop bottleneck, MMDS and LMDS both provide another conduit for broadband services to reach the masses. How quickly (or whether) these technologies move beyond the business market will depend on how rapidly they can reach consumer-friendly prices. Meanwhile, other technologies aim to target consumers from the outset using different delivery mechanisms.

Click here for source article: