fiberopticsonline.com{85A1AB3A-979D-11D3-9A6B-00A0C9C83AFB}&Bucket=HomeFeaturedArticles
Metro WDM Deployment Still Five Years Out, Experts Say
11/10/99 Demand for metro WDM hampered by component cost, limited
demand, and slow data rates in the copper-based access network.
By: Kristin Lewotsky, Photonics Online
For the past year, metropolitan dense wavelength division multiplexing (DWDM)
has been the buzz of the industry. Components and systems vendors have
released a flurry of products, and a number of field trials are underway. According
to market forecaster Pioneer Consulting, the metro DWDM market is slated to
grow from the 1998 figure of $200 million to a whopping $1 billion by 2003.
While attractive for long-haul applications, DWDM deployment in metro networks
presents an elusive business case, where time-division multiplexing (TDM)
upgrades may still be cost effective. Participants in a recent ElectroniCast
seminar (Optical Networking in the Metropolitan Environment; Nice, France),
predict that DWDM is unlikely to enter the metro loop for at least five years, if not
longer.
Prohibitive cost
With cost as the primary barrier to metro DWDM deployment, network designers
initially expected to ease metro DWDM cost by eliminating amplifiers. The
approach has not proven effective. "Without amplifiers, there's a limit to what you
can do," says James Regan, Product Line Manager of Optical Amplifiers for
Nortel Networks.
"The assumption was that power budget is not a restriction," notes Pedro Falcao
Fonseca, director of network R&D at GTS Carrier Systems. "The reality is, it is.
The WDM metro network is made up of passive components. Thus, we do need
amplifiers in the metro space, which drives up cost." The addition of amplifiers
creates the need for wavelength planning, which further increases network
complexity. "We're trying to drop the cost down, but the business case is
currently not there," Fonseca concludes.
"Metro DWDM is caught in the cycle between technology development and
economic parity with alternatives," says Stephen Harbour, vice president of
Communications and Business Systems at Pirelli Cavi. In order for the
technology to take off, it has to provide compelling performance benefits while
remaining cost-competitive. To deliver a successful new system, Harbour says
"you have to get new technology within 20% of cost of existing technology, but
with functional benefits."
Demand and the access network
One compelling functional benefit of DWDM is capacity. But capacity demand in
the metro loop is not high enough to force carriers to incur the increased costs of
optical networking-at least not in the next year or two, according to the
ElectroniCast panel.
Through 2005, the demand for 1.5 Mb/s service in the residential market is
projected to be no more than 15%, with service 2 Mb/s or faster making just over
20% penetration at the end of the forecast period, according to Harbour. Demand
in the small business sector is somewhat more compelling: Through 2005,
projections call for up to 85% of users needing better than 2 Mb/s data rates, with
20% going as high as 45 Mb/s.
A fundamental limitation to the type of burgeoning capacity demand that would
drive metro DWDM deployment is copper-laden access networks, which are
hard-pressed to produce the kind of capacity demand that would force carriers
into DWDM. "We need to get fiber into the local access network before metro
WDM will take off," says Fonseca.
"The biggest challenge [to FTTH] is transceiver cost," says Harbour. "You've got
to have your own laser at each home." Such a requirement is currently
cost-prohibitive. It is possible that wireless technology will overcome at least
some of the problems, increasing data flow without the need for extensive
deployment of a fiber infrastructure. "In 10 years 50% of homes will have FTTH
over radio," Harbour predicts.
Five years out
The panel agrees that the projected time to widespread deployment of metro
WDM is still several years out, perhaps as much as five or ten years. According
to a study by the Data Monitor, voice traffic demand will drop from the 1999 figure
of 330 Pb/mo to 320 Pb/mo. Over the same period, data traffic will surge up from
94 Pb/mo to 6170 Pb/mo, an increase of a factor of 65. "We spent 123 years
building the wrong network," Harbour says, noting that the trend now is toward
data-centric networks.
The panel speculates that average traffic demand may rise to 100 Mb/s in the
next two to three years-not exactly an optical-class data rate. Consequently,
the panel's consensus opinion is that widespread metro DWDM deployment is
still a long way out.
Harbour predicts that metro DWDM efforts will focus on field trials and
development work on management, protection, and provisioning issues.
Widescale deployment awaits improvement of the access loop and networks
costs, consensus in topology, and improvements in protocols, management, and
protection, panelists say.
"Forecasts, especially those based on historic data, regularly underestimate
customer needs," notes Jost Spielvogel, president of transport networks at
Siements ICN, by way of a parting shot.
|
