Mad, here is the article on VDSL.
"Why lay fiber if you can get enough speed out of assets which you
already have in the ground and are payed for??? Wait to see telcos snail
pace on fiber once ADSL / VDSL becomes reality."
Worth more with a V
VDSL spells even higher bandwidth over copper wire--as long as there's
fiber in the loop
MICHAEL SHARIFF
As telephone companies rush to begin delivering asymmetrical digital
subscriber line services, the next generation of equipment aimed at
boosting the bandwidth of existing copper plant promises to deliver 10
times the speed. Joining the digital subscriber line lexicon is very
high speed digital subscriber line technology.
VDSL operates at rates of up to 51.84 Mb/s, compared with the typical
maximum ADSL speed of 6.3 Mb/s (Figure 1). It can be deployed to homes
that are connected via copper to optical network units (ONUs) that, in
turn, are connected to central offices or host digital terminals via
fiber. At a cost of about $1000 per line, VDSL will be economically
competitive with hybrid fiber/coax (HFC) architectures for delivering
broadband services to residential customers.
Initial applications for VDSL are asymmetrical, including
video-on-demand with simulated VCR controls, home shopping and Internet
access, all of which require more downstream than upstream capability to
the end user. Early implementations call for upstream rates of 1.6 Mb/s
to 2.3 Mb/s, although the development of fully symmetrical services is
still possible.
VDSL offers telephone companies an alternative to switched digital video
and cable technologies. Telcos can begin to deploy ADSL, while launching
their long-awaited fiber-to-the-curb (FTTC) platforms. VDSL also
satisfies concerns commonly associated with ADSL about providing
sufficient live video capacity and quality to compete effectively with
HFC networks. Its downstream bandwidth will clearly eliminate any
shortfall on that front.
ADSL is a transition technology that will carry users into the next
century and facilitate migration into VDSL. Because it operates over
longer copper lengths--18,000 feet rather than 4500 feet--ADSL requires
no fiber in the loop.
The cost structures of ADSL and VDSL are more or less the same. While
the infrastructure cost will be higher for VDSL deployment, the
transceiver cost will be much lower--about half that of ADSL. Despite
its slower speed, ADSL is considerably more complex than VDSL because it
must support a wider dynamic range of speeds and distances, resulting in
increased channelization and packet handling requirements.
Board-level component density integration resulting in reduced cost per
port is crucial to successful widespread economical deployment of ADSL
and VDSL technologies. Ensuring smooth migration to new standards is
also important so that service providers' initial investment can be
protected.
Although ADSL is considerably further along in commercial development
than VDSL, all forms of DSL face challenges in the areas of standards,
technology and markets. Despite this, a consortium of Bell companies is
predicting more than 4 million digital subscriber lines to be installed
in the next five or six years--roughly the same growth rate as cable
modems (Figure 2). While initial DSL applications will involve ADSL,
VDSL will begin to appear in three or four years (Figure 3).
Building the network
Because of distance limitations, ONU deployment is vital to the delivery
of higher-bandwidth VDSL services because it extends fiber close to
consumers and is the only reasonable way to carry the traffic that will
be generated by multiple users of VDSL services.
Another key element that will begin to appear in the ADSL environment is
the core asynchronous transfer mode service access multiplexer (SAM),
which is essentially a "magic" box that must aggregate a wide variety of
services from the subscriber side into a highly filled OC-3, OC-12 or
DS-3 pipe interfacing with an ATM backbone switch. For DSL applications,
an SAM would be located in the CO and additional SAMs might be deployed
at remote nodes. Depending on the distance limitations of the specific
DSL technology used, a customer might be served directly from the CO.
When distance limitations mandate, as in a typical VDSL application, the
customer would be served from an ONU, connected by a fiber link to a SAM
(Figure 4).
Vendors have adopted different assumptions about the optimal number of
customers to be served by their remote nodes. Depending on the density
of the area served, estimates range anywhere from 50 to 500 customers.
Defining the SAM is Bellcore document GR-2842, which maps out open, yet
comprehensive, characteristics of the product. Lured by the market
potential, dozens of vendors have participated and competed in requests
for information (RFIs) and requests for proposal (RFPs) issued by
individual service providers and various consortia for the SAM. Most
vendors hope to see refinements to this document that would narrow the
spectrum of services required to keep costs to a minimum, and recent
RFPs issued by service providers seem to have recognized that.
The concept of the SAM has begun to shape the future of digital loop
carrier (DLC) systems, which currently serve 20% to 30% of homes and
businesses, with the remainder fed directly from the CO. In the future,
next generation DLC systems will incorporate SAM-like functionality and
will in effect become hybrid next generation DLC/DSL access multiplexer
(DSLAM) devices. DSLAMs provide some ATM functionality in the loop from
the backbone switch, allowing greater bandwidth management for the last
mile.
Well-designed DSLAM-like products entering the market support ADSL while
also providing a foundation for launching VDSL services. This design
calls for the ability to support high-bandwidth feeds from "any to any"
slot.
Open issues
VDSL's time to market depends on FTTC's time to market--which is at
least three to four years away--and on the resolution of standards and
technical issues.
Five different standards organizations are working on the technology,
including the ADSL Forum, the ATM Forum, the American National Standards
Institute's T1E1.4 group, the European Telecommunications Standards
Institute and the Digital Audio-Visual Council.
These organizations have attempted to minimize conflict by dividing up
the job of producing VDSL standards with minimal overlap. The ADSL Forum
is currently working on network, protocol and architectural aspects of
VDSL. It is relying on the ATM Forum and DAVIC to develop higher-level
protocols and on T1E1.4 and ETSI to develop line code and transceiver
protocols (Table 1).
As with ADSL, a number of possible line codes may be used for VDSL (
Table 2), and it's unclear at this point which of these will emerge as
the industry standard.
Currently, VDSL has only been defined to operate with ATM. Efficient
bandwidth management will require using ATM as a backbone because of its
bandwidth efficiency and its ability to handle the bursty data
characteristics of VDSL.
As DSL technology evolves, so will the encapsulated payload format.
Issues that will shake out over the next few years include whether to
use Ethernet frame or ATM cell transport and whether to extend them to
the CO, the customer premises, or the desktop or set-top.
There are also questions about VDSL's sensitivity to RFI emissions and
its propensity to interfere with amateur radio. Unlike ISDN or ADSL,
short bridged taps and unterminated extensions can cause transmission
problems. Most important, developers must determine the realistic
maximum length of copper wire that VDSL can reliably use to send data at
different rates.
In the realm of customer premises equipment, the most urgent question is
whether to support active termination, which is more expensive but
provides better system management and reliability, or passive
termination, which is cheaper and easier to install and allows several
VDSL modems to be installed at a residence. |
