Nice mention of our dear SFA below!
This story was printed from Inter@ctive Week,
located at zdnet.com.
--------------------------------------------------------------
AT&T Claims Quick Success With LightWire
By Fred Dawson, Inter@ctive Week
December 5, 1999 9:00 PM PT
URL: zdnet.com
In remarkably short order, AT&T says it has made the case internally for a fiber-rich approach to
cable network design that could have major implications for the long-term competitive strength of
cable companies everywhere.
Conventional wisdom in cable engineering circles has long held that deploying fiber-optic
technology deep enough into the cable network to eliminate the need for amplifiers would be a
costly next-generation upgrade that could only be justified once demand for two-way service was
firmly established.
Until that demand materialized, cable operators would have to live with the performance limitations
of established Hybrid Fiber-Coax cable network de signs. With these existing networks, the
number of subscribers on a coaxial link be yond the fiber node might be anywhere from 500 to
1,500, with several amplifiers be tween the fiber node and the end user.
But AT&T Broadband & Internet Services now says it has proven to its own satisfaction that it
can build networks in such a way that fiber nodes will serve no more than 50 to 100 subscribers,
eliminating the need for in-line amplifiers in the process. The cable arm of AT&T also believes it
can build these networks within the cost and service demand parameters of already budgeted
upgrade projects now under way across the country.
AT&T's conviction that this can be done rests on what it has learned during the past few months in
deploying what it calls its LightWire architecture over a 600-mile segment of operating cable plant
in Salt Lake City.
"I think we'll be looking at this approach in a number of locations," says Tony Werner, chief
technology officer at AT&T Broadband (www.attbis.com). "We have a ton of projects already
under way, as well as new ones where we think there's a role for this architecture."
Less is more
AT&T says use of LightWire results in an 80 percent reduction in field electronics, thereby
eliminating key sources of the radio frequency noise that plagues cable systems while cutting back
on the effort that goes into the balancing of signal levels among electronic components in the field.
Moreover, with the LightWire design, the amount of transport capacity available over short lengths
of passive coaxial cable exceeds 1 gigahertz, a big improvement over the 750-megahertz ceiling
imposed by use of amplifiers over longer coax runs.
Werner says that the benefits of "future-proofing" AT&T's cable networks are important
considerations, but he adds that the decision to move forward with LightWire is based on the
tried-and-true cable industry principle that the best network design is the one that does what
needs to be done now at the lowest possible cost.
"At the front end of the project, based on completing a couple of hundred miles of the
construction, we're looking at cost parity with a conventional 600-home node upgrade," Werner
says of the LightWire trial. When ongoing maintenance and operational costs are factored in, the
cost-benefit pendulum swings sharply in LightWire's favor, he adds.
But there are instances where this cost analysis doesn't hold, Werner cautions. "I'm not saying this
architecture makes the most sense for every situation," he says.
One of the key factors affecting potential deployment costs with LightWire is the physical location
of the cable network. "When you have an inordinate amount of underground plant, it may be hard
to get 'LightWire' to prove in, at least in the near term, owing to the amount of fiber you have to
pull," Werner says.
LightWire's cost benefits are likely to hold for upgrades involving networks where the ratio of
aerial to underground cable is "the typical 70-30 split," Werner says. This includes situations
where some of the plant has already been upgraded to two-way system design parameters, where
fiber extends to nodes serving clusters of 600 homes.
Mixing LightWire and existing two-way designs in one cable system is no problem, Werner says.
"We have so much capacity in the 600-home design that we will operate the LightWire segments
as if they were 600-home areas as well for a long time to come," he explains.
The basic LightWire design uses a "multiplexing node" to connect to the last-leg fiber links that
extend to the minifiber nodes, where fiber and coaxial cabling meet in the network. The mux nodes
also are linked via fiber into larger distribution hubs, which in turn are connected via metropolitan
fiber networks to central cable headends.
In the downstream direction, from the headend to the end user, dedicated on-demand signals such
as high-speed data and voice are transmitted over multiple wavelengths from the headend, with
each wavelength consisting of the signals for users connected to a given mux hub. A separate fiber
originating at the hub carries the broadcast cable TV channels, which are combined at the
distribution hubs with the dedicated signals to be delivered over a single wavelength to the mux
node, where they are optically split for distribution to each of the minifiber nodes.
Optical amplifiers at the distribution hubs add sufficient power to the signals to drive the light
through the splitter at the mux node all the way to each of the minifiber nodes, thereby further
minimizing the amount of electronics gear required in the mux node.
There and back
For upstream or return-path traffic from end users to the headend, Dense Wavelength Division
Multiplexing is used. A laser positioned at each mux node combines return-path signals from the
minifiber nodes onto a fiber for transmission back to the distribution hub. At the distribution hub,
the wavelengths are combined and optically amplified for transmission back to the headend.
In its current design, AT&T is using two fibers, one for downstream traffic and the other for
upstream, between the mux hub and the minifiber nodes.
"We might go to one fiber using coarse WDM 'Wave Division Multiplexing'," Werner says,
referring to the two-wavelength combination made possible by operating in the 1,310-nanometer
and 1,550-nm wavelengths.
Along with moving to the LightWire architecture, AT&T has made another key deployment
decision that has important implications for future service expansion as well as current service
requirements. Starting with up grades in its Dallas, Denver and Pittsburgh systems, AT&T will use
a return-path technique based on the digital optical systems used widely in telecommunications,
which the cable industry refers to as digital baseband return.
Initially, AT&T will use the digital baseband return approach, which was pioneered for cable
operators by Scientific-Atlanta (www.sciatl.com), to connect headends to distribution hubs,
according to Tracy Hollings worth, an AT&T Broadband spokeswoman. "The technology could
be extended later and could be used with LightWire as well," she says.
Scientific-Atlanta's digital baseband return system delivers upstream cable signals in uncompressed
digital format, allowing several feeds of upstream signals from the 5-MHz to 40-MHz coaxial
return band to be combined digitally into a 2.5-gigabit-per-second feed using off-the-shelf telecom
lasers. This avoids the more challenging process of using "frequency stacking" for the return
signals, in which each return band is put onto a separate frequency over the return fiber.
Scientific-Atlanta's digital baseband return systems are now designed for use in hubs rather than
nodes, but that will soon change, according to Paul Connolly, vice president of marketing and
network architectures at the supplier. "We'll have product out by the end of the year that will
extend baseband digital 'return' to the node," he says.
When this technique becomes available at the minifiber node or, in the case of LightWire, at the
mux node level, operators will be able to combine up to four feeds from the segmented coax
return paths onto a single baseband laser operating at a prescribed wavelength in the 1,550-nm
window. That signal can then be passed right through the distribution hub via a wavelength
combiner and onto the fiber back to the headend. "This will meet the goal of eliminating the
'distribution' headend," Connolly says, in reference to the racks of electronics now needed to
convert signals at these intersection points.
AT&T's decision to use digital baseband return is leading a wide-scale industry shift in this
direction as cable operators look for ways to maximize efficiency in upstream links, according to
Eric Schweitzer, director of product line management at Harmonic, a cable industry supplier. "This
idea has taken over the industry in the last six weeks," he says, adding that Harmonic is working
on products to meet the demand.
The combination of LightWire design and digital baseband return puts the cable industry on course
to begin looking at what it might do with the additional bandwidth it gains over coaxial cable
through the elimination of in-line amplifiers. The passive cable design will free up more than 150
MHz of bandwidth for new services, including two-way services where some of that bandwidth
can be used for the return signals.
AT&T engineers say that by using digital baseband in both directions for this new service tier, the
company would be able to greatly simplify the operational requirements for two-way services,
possibly even to the point of eliminating the need for cable modems in each home. This is because
the baseband signals at the minifiber nodes could be fed into Ethernet cards and distributed and
accessed over the coaxial distribution area as they would be over any local area network.
AT&T is a long way from making any decisions about this second phase of its LightWire
development project, Werner says. But he makes it clear that the breakthrough to applying the
new architectural techniques for current service requirements at current upgrade costs opens a
window to the future that represents a seamless transition to an ever-more advanced service
package.
"We're in no hurry to figure out what the next steps will be," he says. "But it's nice to know we'll
be able to take them fairly painlessly when the occasion arises." |
