Interesting numbers on costs of installation of some last mile alternatives, about half-way down:
americasnetwork.com
CANARIE migration
Canada is seriously considering becoming the first country to deploy fiber
to the home.
By Annie Lindstrom
Perhaps the best way to meet the bandwidth needs of the
masses is to build a new fiber optic network to each and every
home. To many, that proposition sounds preposterous; to a
small, yet growing number of people, it has the ring of truth
that cannot be denied.
In a day and age when efforts are set on maximizing copper and embedded
plant, Bill St. Arnaud has restored the notion of bringing fiber to the home
(FTTH). St. Arnaud is senior director of network projects for the Canadian
Network for the Advancement of Research Industry and Education Inc.
(CANARIE; Ottawa, Ont., Canada).
St. Arnaud's paper, entitled "Gigabit Internet to Every Canadian Home 2005,"
appeared on CANARIE's Web site earlier this year. Since its appearance,
building a divergent, rather than convergent, third network has become a topic
of serious discussion in Canada and has sparked keen interest in the idea in the
U.S., especially in Washington, D.C.
It is St. Arnaud's contention that digital subscriber line (DSL) technology and
cable modems are fundamentally flawed solutions for providing business and
residential customers with high-bandwidth network access. First, neither cable
modems nor DSL can serve each and every customer who wants them.
Second, they are not future proof. That's because they don't scale to provide
customers with the bandwidth they will need and desire once they get a taste of
their first megabit. In short, St. Arnaud says that service providers might as well
spend the money they are spending to extend DSL's reach on deploying FTTH,
a technology that will last further into the future than either cable modems or
DSL.
"One-third of the world's cable modems are deployed here in Canada. At some
point in the future, DSL and cable modems are going to run out of gas and we
are going to need FTTH. We should start thinking about that day now," St.
Arnaud says.
Not another nightmare
In the U.S., FTTH conjures up images of Time Warner's infamous Orlando trial
and dollar signs spinning quickly down a drain. However, the gigabit Internet to
the home (GITH) network St. Arnaud proposes will cost far, far less, he says.
That's because it will serve only one master — the Internet (at least initially).
Customers would continue to meet their voice and video needs using existing
telco and cableco networks until those services could be migrated cost
effectively onto the GITH network. The other major difference between the
existing networks and the GITH network is that the GITH will not belong to the
service providers themselves. A wholesaler/operator of some kind will provide
competitive open access to the fiber, or wavelengths, that reach into the home
to the service provider of the customer's choice.
St. Arnaud believes that making FTTH a reality likely will require some degree
of government involvement. He envisions municipalities as providers of the
funding, as well as being wholesalers of the access. He points to Canberra,
Australia, where the city is building fiber out to people's homes and opening it
up to service providers who gain access to the network at the central office.
State governments deployed statewide fiber optic networks in the 1990s.
Municipalities have already begun, and will continue, to install citywide networks
well into the new millennium. However, most of the cities that have already
begun such projects, including Spokane, Wash. and Palo Alto, Calif., are
talking about delivering integrated services (voice, video, data and Internet
access) over their networks. St. Arnaud believes they should instead focus their
efforts on providing the only clearly market-driven service to date — Internet
access.
In North America, fiber itself is closer than many people think. Chicago, for
instance, already has three fiber optic networks in place — one for 911, one for
its streetlights and another for its transportation system, St. Arnaud says. But
trials won't be limited to urban areas.
"Some rural communities are in better shape in terms of their ability to organize
themselves and deploy the fiber," he notes.
Canadian government officials and CANARIE (itself a government-funded,
nonprofit organization) could make an announcement regarding GITH pilot
programs later this year or early next year, according to St. Arnaud. The pilot
programs will test some of the concepts spelled out in the discussion paper and
it's likely that they would be launched in each province, he adds.
"With the pilots, we'll work with equipment suppliers to define architectures that
allow competitive equal access to the doorstep in selected communities and
work on proving the business case," St. Arnaud says.
Nowhere to run, nowhere to hide
Although Canada's telephone companies are not 100% enthusiastic about the
GITH effort, they are "not close-minded about it," St. Arnaud says.
"There is recognition among them that maybe DSL is not going to get them all
the way to the goal line. DSL doesn't get to all homes, and if someone helps
them up front with the investment in the fiber, they'd consider simply operating
and managing it in a competitive access situation," he adds.
According to the paper, cost studies are compelling as well. Unlike past FTTH
visions, the GITH network does not have to deliver legacy services. It also
saves on costs because it can aggregate traffic locally via what St. Arnaud has
dubbed a "neighborhood routing puck."
While past studies have set the cost of delivering FTTH at $1,500 to $3,000
per subscriber, GITH is estimated to cost between $900 and $1,000 per
subscriber. That's not much more than the $600 to $900 cost of installing a
cable television (CATV) connection. The GITH cost is actually on par with the
$900 to $1200 cost of installing a residential phone line and less than a hybrid
fiber coax (HFC) connection, which costs between $1200 to $1500 per
subscriber, according to the paper (Figure 1).
"DSL systems have a capital cost of $500 per subscriber, but DSL and cable
modems have a short life cycle due to the rapid obsolescence of the
technology," the paper says. In contrast, the major components of the GITH
system can be amortized over a period of 20 to 30 years.
Although a "considerable amount of research and development" will be needed
to define the appropriate GITH architecture, "wide-scale, economical
deployment of GITH is possible within the next three years," by tailoring
attributes of existing optical Internet backbone networks such as CANARIE's
CA*net 3, St. Arnaud states.
Instead of a gigaPOP, a neighborhood competitive access interconnection point
(NCAIP) could provide end users with simultaneous access to multiple service
providers (Figure 2). There might be competitive NCAIPs and the customer
would have independent connections to different NCAIPs. Dense wavelength
division multiplexing (DWDM) could be augmented by access technologies such
as passive optical networking (PON), shared gigabit Ethernet coax, wireless
access and dedicated FTTH.
Additionally, high-performance routers in gigaPOPs would be replaced with
chip-based routing pucks in the neighborhood service pedestal. The routing
puck is a combination of a Layer 2 switch and a Layer 4 logical switched router
that links to a PON at the customer premises on the line side and to other
routing pucks or the transport network on the network side. It will support
simple, high-speed forwarding at Layer 2 and allow a subscriber to be part of
separate address spaces of different service providers by supporting logical
switched paths between the subscriber and a competitive access service
provider, the paper explains. The paths could map to a single fiber, a
wavelength or a virtual path.
Roadblocks
Just how soon the mass market will need more than a megabit of bandwidth is
difficult to say but there are some applications waiting in the wings for more
bandwidth today. Those applications include digital video image transport,
mega-e-mail attachment transport, DWDM caching and multimedia push and
always-on applications.
Building a third network for purposes of reaching the Internet may sound like a
very good idea on paper, but there are technical and social hurdles to jump
before GITH can become a reality in Canada or the U.S. Those hurdles include
considerable inertia in the form of existing service providers who will see that
network as a serious threat to their futures. Also, there will be a need to gain
regulatory support and drum up local government interest in funding a network
of which it will not be the sole benefactor. Then there's the issue of selecting a
wholesaler and determining how many and which competitive access providers
may gain access to the routing pucks.
At the end of his paper, St. Arnaud points out that the great infrastructure
deployments of the 20th century required a strong partnership between industry
and government. But the GITH network, which is premised on competitive
equal access, will require the defining of a new set of relationships between
industry and government, he says. Building that new bridge would seem to be
the most important piece of architecture that needs to be built and it must be
done well before the first fiber in the GITH network is installed. |
