Ops to Get Auction Shot At High-Frequency Bands

MSOs and incumbent telcos will soon get a shot at bidding
for access to high-frequency wireless-broadband spectrum thanks to a newly formulated
auction policy that could significantly affect the competitive landscape in local access.

The Federal Communications Commission set April 11 as the
starting date for an auction covering the 39-gigahertz band, encompassing 1,400 megahertz
of spectrum between 38.6 GHz and 40 GHz. The FCC plans to follow up at an unspecified date
with another auction covering the 37-GHz band, from 37 GHz to 38.6 GHz.

Cable companies and local-exchange carriers will be free to
bid for licenses in these bands, giving them access to new point-to-multipoint
fixed-wireless technology similar to that being deployed in the LMDS (local multipoint
distribution service) band. Incumbent wireline operators were barred from participating in
that spectrum auction 20 months ago.

Much has changed since the LMDS auctions, with deployments
of advanced systems under way not only at that tier -- which spans 1,300 MHz in the 28-
and 31-GHz bands -- but at 24 GHz and 39 GHz, as well.

With the head start that licensed entities in these bands
have, and with the growing pace of local competition in general, the commission saw no
reason to continue keeping incumbent operators from bidding on the 37- and 39-GHz
spectrum. The same will hold for the 24-GHz tier when it comes up for auction at some
unspecified time next year.

"The commission has been moving in the direction of
flexibility," Wireless Communications Bureau deputy chief Kathleen O'Brien Ham
said. Along with expanding eligibility, the commission has approved use of the spectrum
for virtually any fixed application -- from infrastructure backhaul to Internet access to
multichannel video -- and for mobile applications, as well, she said.

Speaking at a recent symposium on the 39-GHz tier, Ham also
noted that license-holders will have great freedom in the way they divide their holdings.
"If you acquired 1,400 MHz and decided you didn't need 750 MHz, you could sell
it," she said.

Virtually all of the major markets have license-holders in
the 39-GHz band, including WinStar Communications Inc., Advanced Radio Telecom Corp. and
AT&T Corp.'s BizTel unit, according to Jonathan Atkin, principal analyst for
telecommunications services at investment-banking firm Ferris, Baker Watts Inc.

These entities will be aggressive bidders for the spectrum
that hasn't already been allocated under former rules, Atkin noted. "It seems
somewhat doubtful that 39-GHz auctions would present a sufficient platform for an entirely
new business to enter the broadband-services space," he said.

He added that potential bidders should also take into
account the future auctions at 37 GHz and 24 GHz to ensure that they don't get caught
up in a bidding war with 39-GHz incumbents if they can wait for future auctions when bids
might be lower.

Whatever strategy they pursue, bidders should no longer be
in doubt as to the utility of the spectrum, Atkin said. "A lot of the technology risk
in 39 GHz from an investor perspective is limited because WinStar has paved the way with
deployments in this country," he added.

WinStar has begun deploying point-to-multipoint
wireless-broadband systems, but it has been operating point-to-point connections along
with fiber-based networks for several years. The company now holds fixed-wireless licenses
in all of the top 60 markets -- including some in LMDS, as well as 39 GHz -- with spectrum
averaging about 500 MHz per market in the larger regions.

Mirroring the pre-auction licensing scheme at 39 GHz, the
licenses to be auctioned at that band -- and, later, at the 37-GHz band -- will be
apportioned in pairs of 50-MHz spectrum blocks to accommodate two-way transmissions.

In the case of 39 GHz, this will result in the allocation
of 14 100-MHz licenses for each of the 172 economic areas of the country, which are
geographic divisions defined by the Department of Commerce.

Opening these spectrum regions to cable operators and
telcos gives incumbent players and newcomers a quick and relatively cheap way to address
demand for broadband services where the market potential is strongest, Atkin said.

"Estimates of how big the broadband market can get
over the next decade range anywhere from $10 billion to $25 billion, depending on how you
define broadband services," he said. "The small-business sector is the
fastest-growing segment of that opportunity."

This happens to be the sector where 39 GHz is the best fit,
he added. "In the millimeter-wave sector -- where you have 24-, 28- and 38- to 39-GHz
spectrum tiers to work with -- the market opportunity is probably the larger side of small
and medium-sized businesses and the smaller end of major enterprises," he noted.

This fast-growing sector is beyond the reach of most cable
plant, and is not a good market for DSL (digital subscriber line) services because of
limited bandwidth on that platform, Atkin added.

Nor does the typical small-business building environment
work well for MMDS (multichannel multipoint distribution system) and other lower-frequency
ranges where spectrum is limited, he said.

"Our sense is that data in [MMDS and other microwave]
bands is limited to fractional T-1 types of service," he added, "so the market
opportunity for these players is probably the high-end residential user or the small end
of the business application."

Notwithstanding the ideal matchup between the size of the
market opportunity in the small-business category and the applicability of the 39-GHz and
other millimeter-wave frequencies, there's still a lot of doubt about the viability
of these high frequencies for use in broadband services.

The big spectrum-holders -- including Teligent Inc. at 24
GHz and Nextlink Communications Inc. in LMDS, as well as WinStar and ART at 39 GHz -- have
only just begun to deploy new point-to-multipoint systems offering a full slate of
broadband services to business users with the flexibility and efficiency customers demand.

"To us, this market is still changing very
dramatically from month to month," said Doug Gray, chief technology officer of
wireless-broadband-access systems for Lucent Technologies. "Our customers are
considering new applications all of the time, but the technology certainly is
viable."

The key to making good use of the 39-GHz band is to
understand that what might seem to be disadvantages in propagation characteristics can be
turned to advantages with proper planning, Gray said.

For example, to cover an area of 100 square kilometers, it
might take nine transmitter hubs operating at 39 GHz, versus four at the 28-GHz LMDS tier,
assuming that power levels at the transmit and receive ends are roughly equal between the
two systems.

But "more closely spaced hubs can greatly increase
coverage -- which is to say, the number of buildings in line of sight -- and that is a big
cost factor that will tend to offset the higher infrastructure costs at 39 GHz," Gray
said.

Another factor leveling the playing field between lower and
higher millimeter-wave operating regions is antenna-size requirements. Higher frequencies
use smaller antennas for any given level of antenna gain. So an antenna operating at 39
GHz, sized to match the acceptable parameters of an LMDS antenna, will deliver enough gain
to "just about cancel out the excess-path loss" at 39 GHz, Gray said.

Areas with lots of rain can pose problems for 39 GHz versus
the lower-tier spectrums, Gray noted, adding, "One has to be concerned about whether
there are sufficient customers in that small area [of signal propagation] to justify the
costs."

That problem, too, can be offset by the lower costs of
deploying hubs that deliver lower capacity because they serve fewer customers, Gray noted.

"A good metric for evaluating these systems in the
business case is to look at your cost per covered building," he said. "In many
cases, having more closely spaced hubs can decrease the cost per covered building."

The next few months will be crucial to demonstrating how
well point-to-multipoint 39-GHz systems, such as those Lucent and its vendor partners are
supplying to WinStar, work in commercial deployments.

A good early track record would suggest that incumbent
cable companies and telcos would be risking a big chunk of potential business if they
choose to pass up the opportunity to acquire wireless-broadband spectrum.