re: Integrated Access Points: The Next Step for MSOs
Dave Horne, I pointed this one at you because I'd like to hear your (as well as all others') thoughts on this.
From Telecommunications Magazine
Focus on Turbocharging the Last Mile
[Note: the web site has a helpful graphic; text copied below for posterity]
telecommagazine.com
Enjoy, Frank Coluccio
=============
Functionality that's redundant in cable modems, digital set tops and cable
telephones is being moved into one side-of-the-house box.
Rouzbeh Yassini and William Kostka
With U.S. cable modem penetration rising rapidly and millions of
digital TV set-top boxes on order, cable companies are arguably
winning the race to bring broadband connectivity to the home. But
top strategists know the race is not over. Digital subscriber line
(DSL) and broadband wireless (using MMDS, LMDS or other
spectrum), the two most robust challengers to cable and its fire
hose-sized hybrid fiber-coax (HFC) pipelines, are coming on strong.
Companies betting on HFC are casting about aggressively for
technology breakthroughs. Among their current initiatives:
• Enabling cable modems and digital set tops to
incorporate new features and assisting vendors in
producing the increasingly powerful devices;
• Seeking to move functionality that's redundant in cable
modems, digital set tops and cable telephones into one
side-of-the-house box—an integrated home access
point;
• Piecing together solutions to the tasks of provisioning,
security and event-based billing that are rugged enough
to be scalable to a million-user-node network.
The big payoff will begin in 2001-2002, when the integrated devices
start shipping.
Maybe this is a killer app, but its time has not yet come. This market
will take off when voice, data and video are available interactively, on
demand. We can expect:
• Electronic commerce to reach the big-time, eventually
rendering half of today's retail stores nonviable.
• A boom of on-line sales of “electronic products,” with
elimination of the mail delivery service for products and
services ranging from compact discs to college courses.
• A telecommuting explosion, as people realize they can
be more productive in their broadband homes than
commuting to and from offices.
Getting There
In the short term, getting to where this broadband experience takes
hold involves a lot of small, incremental steps to advance the
technology. With cable modems, the recent focus has been on
certifying modems and qualifying cable-system headend equipment as
compliant with what is now called CableLabs® Certified™ Cable
Modems and was known as Data Over Cable Service Interface
Specification (DOCSIS) 1.0. The next phase—transitioning to
DOCSIS 1.1, which supports time-sensitive applications such as
high-quality voice and video—is in midcourse.
Inevitably, future revisions of the DOCSIS spec are intended to take
advantage of further siliconization and slash power requirements from
the 12-15 watts of first-gen modems to 2-3 watts (low enough to be
powered over the network). Furthermore, the specification should
incorporate the lessons learned by operators about best practices for
provisioning and maintaining the network, and measuring and billing
for services. Event-based billing is key.
Other new initiatives for the DOCSIS project include:
• An ongoing advanced physical layer (PHY) project,
seeking ways to squeeze more data into given units of
bandwidth;
• Phasing in universal serial bus (USB) connectivity,
replacing the present 10BaseT Ethernet linkage between
cable modems and PCs—a plug-and-play upgrade that
should speed installations and allow for more
self-provisioning;
• A recently initiated analysis of in-home networking,
looking at vendors' competing approaches: phone line,
power line and wireless.
Why an Access Point Box?
The plan for an integrated access point, or cable access unit, would
move redundant functionality away from cable modems, set-tops and
cable telephone multimedia terminal adapters (MTAs) and into a
shared device.
Home network interface units (NIUs) have been discussed for years.
Previous NIUs have solved only one of the three problems—voice,
data or video. This new access point will handle all three. The access
point box could be located either inside or outside the home. Putting
it outside is favored because it would enable cable companies to
establish a clean demarcation line between their plant and the in-home
network and devices, as telephone companies have been able to do.
In addition to the fact that there would just be one network interface
to the external RF network rather than interfaces in all three devices,
integrated access units have other benefits:
• Instead of all the boxes receiving the entire data
stream coming into the house, the integrated device at
the access point parses the incoming stream and directs
pieces of it over a wired or wireless home network to
their intended computers, set-tops (or digital TVs) and
telephones.
• Powering could be moved to the access point, which
would draw its power either from the external network
or from house current with battery backup.
Participants in two other major initiatives are studying the impact of
an integrated access point on other devices in the home.
The OpenCable project is defining specs and nurturing vendors for
digi- tal set-top boxes compliant with the OpenCable specification.
These boxes combine analog video and compressed digital video
delivery and could also include a DOCSIS cable modem, permitting
manipulation of electronic program guides (EPGs) as well as
TV-based Web browsing, e-mail access and other activities. Major
cable operators have ordered millions of OpenCable-compliant
set-tops.
Deploying digital set-tops now will prove to be a good move even if
much of their functionality is later moved into access point boxes. In
particular, they're the fastest way to get digital television, EPGs and
Web-based interactivity into TV sets. The name of the game now is
to keep subscribers.
The PacketCable project is writing a series of specifications for a
more generic form of packet-based, digital connectivity,
encompassing services like IP telephony and videoconferencing.
Digital integration may have its limits. Specifically, cable engineers
think the day is far off when their primary products—broadcast and
on-demand video streams—become part of an IP data stream. The
FCC regulations requiring cable systems to deliver analog video to
existing cable-ready TVs and analog set-tops is a barrier to
IP-everything.
There's also the matter of bandwidth efficiency. On today's cable
systems, both broadcast video and DOCSIS data are embedded in
MPEG for downstream transmission. Broadcast MPEG was
designed specifically for cable networks and seems to do a pretty
good job. Should the digital video that's now running on the raw
transport stream be embedded into IP so it can run in DOCSIS,
which then runs on a native stream? The big question is whether that's
a wise use of bandwidth.
The video on demand being tested by several cable operators is
digital video from servers at local headends, so it's not going across
anyone's IP backbone. If multiple customers are going to be served
with different video streams, all at high speed, the servers must be
close to the location of the TV sets. Otherwise, the demand for bits
would rapidly eat up any IP backbone.
Similarly, current cable phone services carry voice traffic on
n-by-64-kbps digital voice channels rather than in IP data streams.
While some operators are rushing to embrace IP telephony, others
are in no hurry to abandon cable phone, which has become a
profitable business. However, future telephony deployments by major
operators such as AT&T and Comcast are expected to be IP-based.
The Consumer Experience
ILECs and CLECs will also need integrated digital offerings and
access point devices. They believe most consumers will be content to
do business with one broadband service provider. Consumers will
mostly keep using their existing TVs, PCs and telephones and will be
largely unaware of the access point box. The true broadband
experience requires that the end user not have to understand or deal
with the technology behind it.
One of the biggest selling points to consumers will be lower prices.
There are economies of scale from providing voice, video, and data
over one network. A customer doesn't care about the home
infrastructure. He'll use his TV as he always did. He'll call on his
phone as he always did. Having a common infrastructure will enable
services, much like Microsoft was able to “standardize” Windows to
the point that everybody makes software for it. Once DOCSIS
connectivity is there, everyone will write applications to run on top of
it.
Rouzbeh Yassini is the father of the cable modem. CEO of YAS
Corp., Yassini is involved with CableLabs® and the modem
initiative as a senior executive consultant. He is the former CEO
of LANcity, a cable modem pioneer.
William Kostka is a project director for broadband Internet
access activities at CableLabs®, focusing on the DOCSIS
physical layer and test issues of DOCSIS equipment certification
and qualification. Previously, he was a member of the
Hewlett-Packard QuickBurst cable modem and Kayak digital
set-top converter projects.
RSNo. 313 |
