|
Broadband Access
Copper holds a tight bond
Emerging techniques provide interim step toward the fiber end game
by Sean Buckley
Just like Polident for dentures, a service provider’s bond with copper
continues to be tight.
Now, through the advent of copper bonding techniques — including
bonded T1 (Overture Networks and Turin Networks), DSL (RAD),
2-BASETL (Actelis and Hatteras) and multiple input, multiple output/discrete multitone
(Aktino) — the service provider’s bond with copper is becoming
even tighter.
Service providers are finding the copper line can be used not only to
deliver Ethernet services but also to lower DSLAM backhaul and wireless
backhaul costs.
Insight Research confirms such an opportunity is on the rise.
In its report Telecommunications and Bonded Copper: A Solution for
Carrier Ethernet, Wireless Backhaul, and DSL Backhaul 2008-2012,
Insight Research forecasts the revenue from Ethernet over copper
services will spike from US$146 million in 2008 to about $1.7 billion in
2012 (see Figure 1).
Insight believes these forms of copper bonding can serve as a sound
interim solution when a service provider can’t make a business case to
deploy fiber.
Copper bonding enables wireless operators to maximize the abundance of
copper already present at cell sites. “A return on investment has to
show positive results in order to pull fiber to any location, and if you
can’t meet that criterion you’re going to continue to depend on copper,”
said Bob Rosenberg, President of Insight Research. “The pair bonding
solutions present a viable route to handle traffic until a fiber investment
can be justified.”
Rosenberg added bonded copper will help service providers expand
consumer and enterprise service rollouts. “This is a way to defer
investment and get more bang out of the buck from the current plant,”
he said.
Closing the enterprise gap
It would be hard to argue that, like peanut butter and jelly, Ethernet
service and fiber is a good match.
When large carriers (AT&T and Verizon) and upstarts (Yipes, which was recently acquired by Reliance Communications) began rolling out Ethernet service, the focus was on fiber-based
offerings. Looking to get a big payback on their respective Ethernet
investments, these operators targeted tall shiny buildings with customers
willing to pay big bucks for premium Ethernet services.
This drive, however, left smaller businesses and even remote facilities of
larger companies with few options other than DSL, T1 or higher-priced
TDM circuits. Currently, only about 14 percent of buildings have fiber,
leaving 86 percent without access to fiber or Ethernet service.
Still, bandwidth hunger creates an opportunity for service providers to fill
the fiber gap with Ethernet over copper.
“Today, branch locations are served by copper; there’s no fiber
[trenching],” says Rich Power, vice president of Insight
Research. “These customers are looking for solutions that can provide
high-speed connections. There’s a tremendous pent up demand to
connect these smaller locations that are on copper.” (See:
Audiocast: Bonding with copper)
After getting a shot in the arm when the former BellSouth, now AT&T,
started to roll out Ethernet-over-copper service, other competitive
service providers and incumbents have been aggressively rolling it out.
Take Cavalier Telephone. Cavalier has an extensive fiber network that
consists of 8,000 intercity and 3,000 metro route miles of fiber. The
advent of Ethernet over copper enabled it to expand the reach of its
Ethernet Express offerings from 3 Mbps to 10 Mbps in its seven-state
territory.
Similarly, Time Warner Telecom needed a means to expand its Ethernet
footprint where it did not have available fiber. Even though TWT can provide fiber-based Ethernet services to 8,100 buildings in the US, Mike Rouleau, senior vice president of business
development said, “it still not enough to reach out to all locations that a
customer wants to take advantage of Ethernet services.”
Rouleau adds that branch offices are ripe for Ethernet over copper. To
expand its Ethernet service to a larger audience, TWT deployed Overture
Networks’ ISG platform.
“We’re seeing this huge trend where customers have built out their
backbones and are connecting their primary locations, but now they
want to extend out to branch offices,” he said. “They want to be able to
take advantage of Ethernet technology for its scalability and plug-and-
play nature and leverage that back into the backbone, so that’s where
Ethernet over really copper comes into play.” (See the exclusive Webinar+Podprofiles — Ethernet over
Copper: Filling the Fiber Gap, available on demand Feb. 13, 2008)
Stretching the last mile
In the United States, the last-mile access network continues to be a
tale of two approaches: fiber to the premises/fiber to the home and a
hybrid copper solution via fiber to the node or fiber to the curb with
either ADSL2+ or VDSL2 as the link into the customer home.
Unsurprisingly, all eyes are focused on AT&T and Verizon’s next-gen
broadband rollouts.
While Verizon continues to expand its DSL service in targeted areas, its
main focus is on its FTTH-based FiOS service, which currently passes
more than 1 million homes.
Alternatively, AT&T is sticking to its FTTN guns for existing markets and
FTTP for new Greenfield deployments. CFO Rick Lindner said in AT&T’s 4Q07 earnings call that copper bonding “will not quite double but significantly increase” the amount of
bandwidth delivered to the home. Initially slated to start in 2007, AT&T’s pair bonding initiative to support enhanced HDTV services won’t begin until late 2008.
Insight’s Power believes the adoption of bonded copper or fiber really
depends on some fundamental factors. “The population density, the
carriers planning horizon, and the consumers appetite for broadband will
drive the decision in the long-term,” he said.
Outside of AT&T and Verizon, the question is what will other ILECs such
as EMBARQ do?
After reaching its 1 millionth broadband customer in late 2006, EMBARQ
has continued to use a combination of fiber and traditional copper. In
late 2007, EMBARQ launched a 10 Mbps ADSL2+ service and in 2008 it
will launch an expanded reach DSL offering targeted at extending service
to its rural customer base.
Overall, Jim Hansen, senior vice president of network services for
EMBARQ, is upbeat about various techniques proposed by the vendor
community (i.e., Dynamic Spectrum Management and pair bonding) to
use their existing copper plant.
“We see these as ways to extend, or what we call stretching the
copper,” he said. “Because of the way the network is built, pair bonding
is an attractive [way] to use an asset that’s sitting there fallow. These
are the types of things we’re looking for over the next two to three
years to get 25 Mbps to over half our DSL customers.” (See:
Audiocast:EMBARQ bulks up on broadband)
But copper bonding is not just about services alone. In areas where it is prohibitive to build out fiber, service providers can leverage their existing copper infrastructures to backhaul DSLAM traffic.
“Consumers are demanding an increased amount of broadband capacity,”
Power said. “Particularly in rural geographies, service providers that have
not installed fiber to these backhaul locations will have an opportunity to
get higher speed circuits using Ethernet over copper.”
The wireless backhaul equation
While the wireless network’s so-called killer application will continue to
be voice, the increased adoption of new data and multimedia video
services in the 3G and 4G networks will place higher speed requirements
on the backhaul network.
Traditionally, wireless operators have used either T1/E1 copper circuits
or microwave to backhaul wireless traffic from the base station into the
cellular network.
Given the fact thousands of cell copper connects sites, wireless
operators potentially could bond those existing pairs to get higher
bandwidth.
Of course, Ethernet over copper will face the initial basic operational
challenges.
“Anytime you change to a new technology like the introduction of
Ethernet, there are a lot of concerns,” said Power. “Will we be able to
maintain the same level of performance? So there’s a lot of testing and
evaluation going on at this point.”
Even so, Power believes Ethernet over copper could be a breaking point
to get Ethernet in the backhaul network. “The fact they have so many
cell sites that don’t have fiber is going to force them to consider
Ethernet over copper as a logical solution,” he said.
Although regional operators such as Alaskan-based Mantanuska
Telephone Association and Canadian-based MTS Allstream initially
embraced Ethernet over copper, large wireless operators are also tuning
into its utility for wireless backhaul.
Sprint, which incorporates a mixture of traditional LEC T1 circuits,
microwave and fiber in its wireless backhaul arsenal, believes Ethernet
over copper is a viable near-term solution.
“For all 3G cell site backhaul, we think there’s an interim niche, and an
interim could be a long time,” said Craig Cowden, vice president of cable
operations and interconnected access solutions at Sprint. “We are
working with providers right now and we think we’ll be able to test the
viability of a 10-meg, Ethernet-over-copper solution. We’re very
interested in it as an interim technology, and we will go to fiber as a long
term solution, but long term can be just that.”
Despite the promise Ethernet over copper holds for wireless backhaul,
not everyone agrees it’s the most optimal approach.
Selling wholesale capacity to mainly Tier 2 and Tier 3 wireless carriers,
Southern Light, is quick to acknowledge wireless operators are definitely
interested in using Ethernet. (See:
Southern Light supercharges Ethernet)
And while he can see the argument to use Ethernet over copper for
backhaul, Eric Daniels, Southern Light’s COO, said he’s sticking with
fiber.
“Fiber’s resiliency means we don’t have to do the same amount of
maintenance per foot, which means I don’t need the same number of
people and thus can spend more time doing installs and less on repairs,
and can pay the people we have better,” he said. “I think we would
need to quadruple the size of our operations department in all divisions if
we operated an EoC network, from the NOC to outside plant to field
techs.”
|