 |
|
 |
By Jonathan Collins
Originally Published: June 5, 2001
By: Tele.com
Free-space optics, a failure in the late 1980s, tries for
a comeback in the new millennium
Here's a blast from the past: free-space optics. The
technology, also known as laser networking, was thought to
be consigned to the dustbin of telecom history when it
failed to deliver in its first commercial incarnation more
than 10 years ago. What's old is new again, however, and
today a group of established vendors and startups are
offering this old chestnut to Internet service providers
(ISPs) as a way to roll out reliable high-speed, last-mile
connections quickly and at a fraction of the cost of laying
fiber.
Startups such as AirFiber Inc. (San Diego), fSONA
Communications Corp. (Richmond, Canada) and Terabeam
Networks Inc. (Seattle), as well as Lucent Technologies
Inc., are trying to convince carriers that they've got the
technology right this time. But that can be a hard sell.
"It's a technology area that is tainted," says Jim Dunn,
president, CEO and cofounder of AirFiber, which launched its
first free-space optical products in April. The earlier and
far more expensive generation of products promised too much,
he says, in reliability and distance.
Free-space optical equipment uses lasers to transmit
high-speed data over the last mile and connect buildings and
businesses to the fiber backbone networks. The technology
uses unlicensed spectrum and promises transfer rates in the
gigabit range. Lucent is talking about the possibility of
40-Gbit/s connections.
Both the landscape and equipment have changed since the
early failures. Rising demand for high-bandwidth last-mile
connections, combined with the drop in price and size of
free-optic equipment, has rekindled interest in the
technology. "Free-space optics satisfy a very urgent need
for bandwidth in the last mile," says Bettina Tratz-Ryan, a
senior analyst at Gartner Group Inc. (Stamford, Conn.),
adding that it offers tenfold savings compared to fiber
deployment. Proponents say laser networking is cheaper than
digital subscriber line (DSL) and fixed wireless in addition
to fiber. "Free-space optical is the cheapest cost per bit
available, and we can prove it," says Theresa Carbonneau,
president and CEO of fSONA. Her company has equipment in
trials with competitive local service providers and ISPs in
the United States and is slated to ship its first products
next month.
This promised low cost could completely change the business
model for the deployment of high-bandwidth last-mile
connections, says Dunn. He says an AirFiber 622-Mbit/s link,
which goes for about $20,000, creates far more flexibility
than a $100,000 fiber connection. The carrier can deploy
free-space networks and then begin selling capacity to
customers or other carriers. The high cost of deploying
fiber means that connections to the largest clients are
built only after they sign on the dotted line.
Regardless of the savings, carriers and ISPs will have to be
convinced that the new generation of free-space optical
equipment has overcome its earlier weaknesses. Like fixed
wireless, free-space optics requires direct line-of-sight
access between the transmitting and receiving nodes. This
means all kinds of factors can break the connection, such as
trees and new buildings. The lasers can even move out of
alignment when the skyscrapers to which the nodes are fixed
sway slightly in high winds. Other weather conditions can
also hamper performance, with fog posing by far the most
significant threat.
The vendors behind the latest free-space generation are
quick to claim that these problems have been solved. By
connecting nodes in a mesh format, each node has an
alternate route in case of temporary blockages. Nodes are
moved closer together in places where weather is an issue.
AirFiber says it can place nodes as far as 450 meters apart
in predominantly dry conditions. As for permanent blockages,
such as new buildings between two nodes, nodes can easily be
moved and meshes reconfigured. But reliability shouldn't be
an issue, says Carbonneau. "Everyone is wary of the
reliability and quotes the 99.999 percent uptime as a
target. But carriers understand that is a figure for network
uptime, not the uptime of a single link. Any network setup
should allow for a parachute of some sort," she says.
One such parachute is fSONA's beta test with a carrier in
Vancouver, Canada. The carrier is using the free-space optic
link as a return link for its Synchronous Optical Network
(Sonet) ring. "If there is a problem with the free-optic
link, the Sonet ring becomes bidirectional and carries the
traffic," says Carbonneau. "There is no network
downtime."
While both AirFiber and fSONA rely on rooftop placement for
their nodes, Terabeam has developed a multipoint access
technology using small, satellite-like dishes that can be
fixed indoors next to windows. The aim is to build a
metropolitan network similar to a cellular radio network.
The metropolitan area would have a set of optical hubs that
would beam backbone signals to customer units. One optical
"cell" can serve dozens of individual customers with up to 1
Gbit/s of capacity. Terabeam has already won backing from
Lucent in the form of a $450 million joint venture. One
thing Lucent is getting for its money is sole reseller
rights for the company's equipment. Meanwhile, Nortel
Networks Inc. has signed up to resell AirFiber's products.
Analysts say the presence of Lucent and Nortel in the
free-space market is significant. "That both Lucent and
Nortel have opted to sell free-space optic equipment, even
though they have wireless local loop products, is a
validation of the free-space market," says Gartner Group's
Tratz-Ryan.
Aside from Terabeam's service provider ambitions, the
free-space optic vendors are all targeting carriers,
competitive providers and ISPs looking to quickly deploy
high-bandwidth last-mile connections. This is a huge
potential market. Because the spectrum is unregulated around
the world, the target market is global, not just domestic.
Japan's DDI Corp. (Tokyo) has deployed a test network using
AirFiber's gear, and Broadband Optical Access (Madrid), a
European-based carrier for metropolitan networks, is looking
to roll out AirFiber's OptiMesh network if current trials in
Madrid prove to be a success.
But despite such enthusiasm, the new technology must prove
its value in far more beta tests before the promise of the
new generation removes the doubts caused by the last.
|
 |
