Twisting Pairs:Getting the Kinks Out of Video Over DSL

Posted: 11/2000

Twisting Pairs:Getting the Kinks Out of
Video Over DSL
By Aldo Morri

Let’s face it, carriers still don’t have the bugs out of DSL, especially when
it comes to figuring out how to deliver video over these upgraded, twisted-pair
copper wires.

Nonetheless, as companies attract a critical mass of customers for DSL
Internet access, they also begin to leverage DSL for video services, as they
always promised they would.

High-speed Internet access deployment is moving relatively smoothly.
According to The Strategis Group (www.strategisgroup.com
) in Washington, D.C., more than 2 million U.S. residents will subscribe to DSL
by the end of 2001, and more than 11 million by 2004. Cable company competition
remains strong, however, and high-speed cable modems likely will have at least
as many subscribers as DSL.

To compete, telcos have turned their attention to video. At present, all the
former Bell companies (as well as many incumbent European and Asian telcos) are
deploying or testing video services and so-called "gateway" products,
as they prepare for a market they think will experience triple-digit annual
growth rate through 2003.

Some experts forecast that from 12 to 36 months from now, video content
delivery over DSL in the United States will be at least as widespread as having
a DSL line is now. Several million homes may be using DSL video services within
the next three years. Videoconferencing, which only requires 384 to 500
kilobytes of bandwidth to be commercially viable, will precede video
applications such as movie downloads that require more than 3 megabytes of
bandwidth.

"For the first time, cable has a real challenger," says Steve
Hensley, vice president of engineering, CTO, and co-founder of New Edge Networks
(www.newedgenetworks.com)

Well, maybe not just yet. The industry, devoted to giving cable a run for its
money, is still in its infancy. Operators are spending more of their time
figuring out how to transport VoDSL rather than video.

According to Jason Marcheck, broadband analyst with The Strategis Group, many
issues still need to be resolved before video over DSL becomes commercially
widespread.

"There are vendors out there working on solutions, but the rollout and
marketing push for the services will take a while," says Marcheck.

Whole Lot of Shakin’ Goin’ On

The broadband industry still expects a whole lot of shakeout to occur, along
with technology experimentation. We still have several different kinds of
companies delivering DSL under different architectures, and not all products are
created equal. For starters, some DSLAMs in COs around the country have
significantly impaired capabilities compared to newer ones that are coming out
now. Significantly more is expected from the newest generation of DSLAMs that
will surely appear within the next 12 to 18 months.

"It’s like individual modems," says Hensley. People thought that
when 9,600-baud modems were invented, that we had reached a theoretical maximum
speed for dial-up modems. Six months later, we reached 14,400-baud modems, and
28,000 baud were reached soon thereafter. The rest, of course, is history.

"It’s just a matter of time before these DSLAMs get better. There are a
lot of new things coming up," says Hensley.

Also, DSL comes in several different flavors, and when it comes to the
delivery of video services, the differences make quite a difference. According
to Scott Smith, CEO of broadband CLEC provider MetStream Communications Inc.,
video really needs to be delivered at speeds of at least 3.7mbps over DSL to
function optimally.

"3.7 megabits is the magic number," says Smith.

That leaves symmetric DSL (SDSL) out of the picture for video delivery,
because it maxes out at about 1.5mbps. Asymmetric DSL (ADSL) can deliver video
at speeds up to 8mbps, but only if the delivery point–an office or home–is
within 5,000 feet of the CO. Otherwise, ADSL quickly disintegrates to 2
megabytes per second at 10,000 feet from the CO.

So, most DSL only works very well for high-speed Internet access and a second
phone line, but not for video. That’s why MetStream opted for the most expensive
DSL option, very high data rate DSL, or VDSL.

"Our architecture may cost slightly more, but we will be able to compete
because we have a technology and an architecture that guarantees quality of
service," says Smith.

Computerized TVs or Televised Computers?

Most competitive operators will tell you the main obstacle to a profitable
business model in delivering video is the intransigence of local exchange
incumbents. By accident, or design, these companies put up obstacles to doing
things quickly, many CLECs claim.

But, another problem that no one denies lies in the area of uncertain
consumer demand for video over DSL. More providers seem to conclude that
consumers want such services.

"The more we open up the pipe, the more we make it possible for people
to traverse the expanses of the Internet, the more demand for applications there
will be. It is that virtuous cycle that everyone is talking about," says
Sal Cinquegrani, spokesperson for New Edge Networks.

The cable industry says everything will be delivered via TV, while the
telephony industry makes the case that everything will move to the computer. The
truth is somewhere in the middle and TV and computer will co-exist.

"It’s going to be difficult to get people and businesses to use TVs to
do certain things, and it will be unlikely that you will get people to use
computers to watch TV," says Hensley.

One impediment to using computers as TVs is screen size. Large display
technology is expensive, and customers are not likely to do serious computing on
a 60-inch screen.

On the other hand, consumers might want to download large video applications,
such as movies–as millions are now doing with music at websites such as Napster
Inc.’s (www.napster.com)-and
playing them on their televisions with a device like today’s MP3 for music.
Alternatively, such applications might be networked to the TV, or put into some
memory module that might be used on a laptop on a plane flight.

Meanwhile, everyone is waiting for affordable and reliable one-to-one video
applications such as videoconferencing and training applications.

And, of course, some new invention could come along that changes everything.

"Gateways" Provide the Missing Content

Along with the advent of high-bandwidth technology, a whole new breed of
content company has been created. Often referred to as "gateway"
companies, or "aggregators," the main difference between these
companies and old-fashioned television networks such as NBC and CBS is they
function as cybernetworks, providing multiple channels of information. They have
some of their own advertising, and negotiate rights to access content of all
kinds, from news to movies.

"Managing over 1,500 content relationships is amazingly difficult,"
says New Edge Networks’ Hensley.

In today’s environment, these content gateway organizations, such as iBEAM
Broadcasting Inc. (www.ibeam.com)based
in Sunnyvale, Calif., provide a very important role that leaves engineers with
the communications companies to focus on provisioning and technology issues.
But, because these gateway companies provide their services via satellite, they
have even helped solve some complex technology issues.

The Internet has its strengths and weaknesses, as we have known for some
time. Its most significant strength today is that nearly everyone is now
connected to it. But the major weaknesses are QoS and reliability. Added to
this, DSL is only a "last mile" technology.

It is a fatter pipe between a telephone company’s CO and the home–no more
and no less. It doesn’t solve the issue of getting packets of information across
the country. That’s what the Internet is supposed to do, but, the Internet is
horrible as a delivery vehicle for video.

It is somewhat ironic, because DSL was intended to make the Internet blossom,
but to ensure a quality video service, most DSL providers have to go around the
Internet, be it over proprietary fiber or via satellite.

DSL providers are being convinced the latter method makes more sense.
Companies such as iBEAM partner with DSL carriers to deliver content via a
satellite dish. Servers are placed in the DSLAMs that are served by an iBEAM
satellite connection. The DSL providers can then simulcast content to their
customer base. These satellite-enabled gateways thus help DSL providers to
bypass the Internet. What’s more, iBEAM and others, such as CoolCast (www.coolcast.com)
deliver the content to carriers for free. It’s the content companies that pay
iBEAM to "broadcast" their content.

"The option to our service [for carriers] is to create multiple data
centers that are fed terrestrially," says John Fiske, iBEAM’s entertainment
division director of vertical marketing. "That would be very
expensive."

Larger network providers might be looking to purchase such content delivery
organizations. Companies such as New Edge are not likely to get in this game. It
probably makes more sense for companies such as this to buy one of their
technology vendors. For larger players–some of whom are already mixing networks
and content ownership, such as Time Warner Inc. (www.timewarner.com
) -it is a different story.

But no fears should arise about too much consolidation in this area.

"The multimedia broadband industry is too huge for anyone to
dominate," says Hensley.

Aldo Morri is a free-lance writer based in Washington, D.C. He can be
reached at [email protected].