Another nice article MUST READ with ASND ref. at NWFUSION
nwfusion.com
"ATM is an excellent technology for carriers
that have to structure deployment strategies
in an environment with constrained
bandwidth," says Lew Wilks, president of
Qwest Communications International's
business markets unit in Denver. However,
while Qwest's original intent was to build a
"pure IP" network, market realities - ATM
dominance on carrier backbones and the
relative immaturity of IP switching
technology - forced the carrier to implement
ATM. (Playing into ASND's strength) Bucky what do you think about this?
Qwest has spent $2.6 billion building out a
fiber infrastructure along highly protected
railroad right-of-ways, where buried cable is
least likely to get damaged by backhoes and
tree roots. The company lays two side-by-side
conduits - one with 96 fiber strands and one
entirely empty, to accommodate future fiber
technologies.
"Some of our competitors have only one fiber
on some legs of their networks, while 24 is
the least we have anywhere," Wilks says.
"Rate arbitrage will go away over time, so
margins are what you need to survive, and we
have a very low-cost infrastructure."
Founded by executives from competitive local
access pioneer MFS, Omaha-based Level 3
Communications is approaching end-to-end
services from the opposite direction. The
company has installed local fiber loops in 50
U.S. cities and another 20 or so abroad, and is
connecting the dots with a long-distance fiber
network.
"Voice will eventually be just another
application on an IP network, and ours is the
first network totally optimized around IP
traffic," says Ron Vidal, senior vice president
in charge of new ventures for Level 3. "We
have no circuit switches at all."
Level 3 acquired Xcom Technologies earlier
this year and consequently now owns
technology for connecting IP networks to the
Signaling System 7 call signaling
infrastructure. That will help Level 3 deliver
call control, which has been one of the
missing pieces in IP telephony.
Williams Communications in Tulsa, Okla., is
another one of the young turks starting with a
clean slate and deploying huge amounts of
fiber around the country. The company is
building out its network with 144-fiber cable
and deploying OC192 with dense
wave-division multiplexing on each fiber. The
lack of QoS for IP convinced the carrier ATM
was the only way to go.
Williams eliminated cross-connect equipment
by migrating to GX 550 ATM switches from
Ascend Communications. The company is
also migrating SONET into the ATM switches
and thus avoiding SONET terminal equipment.
(It will be nice if other upstart CLECs start doing the same)
"So we're saving 50% to 80% on hardware
costs," says Wayne Price, manager of
technology development for Williams'
network division.
Some experts doubt that over-provisioning
alone can guarantee business-quality voice in
the long run. The demonstrations sound great
now, but the networks don't yet have any
traffic to speak of.
Nevertheless, all that buried fiber is clearly a
precious resource.
"The cost of switching has gone down by
orders of magnitude, while the cost of
right-of-ways has not changed that much,"
says Hal Varian, dean of the School of
Information Management and Systems at the
University of California, Berkeley. "Fiber in
the ground is like money in the bank."
And the bandwidth bandits are using it to buy
market share. While this is just a short-term
strategy, large corporate customers can take
full advantage of it. "There's only an upside
for enterprises," Soper says. "They are
getting accelerated services at great rates."
ILECS: Eating their young
Clean slates are an advantage in a lot of
ways, but they can also imply lack of
experience. Here the scales are weighted
heavily in favor of the incumbents that have
been deploying and supporting huge networks
for a long time.
"The traditional carriers haven't been the
movers, but they can certainly be very
effective followers," says Neville O'Reilly,
director of enterprise consulting for
TeleChoice in North Brunswick, N.J.
The biggest convergence challenges are at
the edge of the public network, in traditional
ILEC territory. The ILECs are in a difficult
situation, with $20 billion to $30 billion sunk
in circuit-switching equipment on
depreciation schedules of up to 20 years.
They want to protect these investments and
avoid cannibalizing their traditional voice
business.
"If their inaction creates a vacuum,
alternatives will come into play," says Joe
Firmage, chairman and CEO of USWeb, a
Santa Clara, Calif., company that offers
various Web services through a network of
affiliates. "They can either eat their own
young or watch competitors eat them."
Now ILECs seem intent on swallowing up
each other. Bell Atlantic just one-upped SBC
Communications' Humpty Dumpty act of
gluing the RBOCs back together again by
announcing plans to merge with GTE. The
combined $53 billion entity would surpass
AT&T as the largest U.S. carrier.
Many regard the proposed merger with
dismay. GTE has been the wild card in the
traditional ILEC camp. While the RBOCs are
awaiting the regulatory go-ahead before they
can get into the long-distance market, GTE
has been under no restraints.
The company is an ILEC in 28 states from
Hawaii to Florida - albeit largely in secondary
and tertiary markets. Additionally, it bought
24 strands of fiber from Qwest to start
deploying a long-distance network.
GTE expects to be the biggest provider of
ADSL services by year-end. IP telephony trials
have all been internal so far, but the company
plans to roll out commercial IP fax services in
the fourth quarter.
The Bell Atlantic/GTE merger will take at
least 12 to 18 months to complete and may
never go through. Besides significant
regulatory hurdles, the two companies have
divergent management perspectives, and
GTE's shareholders will have to approve a
deal that has the market pricing their stock
at less than it was before the merger was
announced.
Meanwhile, Bell Atlantic is venturing into
convergence waters with a recently formed
subsidiary called Bell Atlantic Data Solutions
Group (BADSG). The idea is to have a
separate, unregulated entity that is a lot more
nimble than the huge parent RBOC and that
can compete on more of a level playing field
with companies such as Qwest, Level 3, and
Williams.
BADSG is in charge of building the new
packet-switched long-distance network Bell
Atlantic announced in June. First-stage
deployment will connect hubs in Boston, New
York, Philadelphia and Washington, with
plans to extend the network throughout Bell
Atlantic's 13-state region, across the country
and around the world. Commercial service
delivery could begin as early as January
1999.
"We will be spending $400 million over the
next three or four years just in the Bell
Atlantic footprint," says Herb Osher, vice
president of marketing for Bell Atlantic
Network Integration, now a part of BADSG.
The ATM-based network will initially be used
mainly for business data services, including
managed IP networks. Because parent
company Bell Atlantic is still prohibited from
offering long-distance voice in its own region,
deployment of converged voice services on
the new network is "contingent on regulatory
relief," Osher says.
US WEST is approaching convergence from
the opposite end. Threatened on the
local-access front by cable-TV companies,
the Denver-based RBOC is deploying
integrated dial tone and "Web tone" services
for enhanced local access. The company has
launched ADSL services in 40 markets across
Arizona, Colorado, Idaho, Iowa, Minnesota,
Montana, Nebraska, North Dakota, Oregon,
South Dakota, Utah, Washington and
Wyoming.
"Using second-generation DSL technology, we
may be able to deliver DSL services over 60%
to 70% of our loops," says John Charters, vice
president of Internet services and application
development at US WEST in Denver.
Indeed, while other telcos focus on laying
fiber, the RBOCs and GTE are effecting a
copper renaissance. SBC is focusing on the
residential market as it successively glues
Baby Bells back together. SBC's Pacific Bell
subsidiary has been particularly aggressive
about rolling out DSL services.
According to Charters, these developments
disprove claims that the ILECs don't get it
when it comes to convergence. "We do get it;
we're just more pragmatic in our approach,"
Charter insists.
The question is whether the market will wait
for them.
"I want to see one of the ILECs step up to the
plate and tell Wall Street, 'We're going to take
some losses while we make major
investments in a converged network,'"
Firmage says. "It would be enormously
painful, but they have the cash to do it with
and the runway to take off from - if only they
have the vision and the courage to proceed." |
