As the decade ticks toward its final fortnight, its biggest
promise -- "the information superhighway" -- has come down to this: a very, very
wide one-lane road.
The fin de siècle vision of universal digital
pipelines is being hyped by "networks" you've never heard of, although you can
get a glimpse of them when they make their cable-industry debuts at this week's Western
These newcomers -- companies such as Chromatis Networks and
Redback Networks (with its newly acquired Siara Systems subsidiary) -- are moving along
the same, very broadband lines that Cisco, Nortel, Lucent and others have already staked
Significantly, the metropolitan-area networking espoused by
these companies is surfacing at a time of cable-industry consolidation and clustering --
ideal for the architecture built around a huge pipeline.
For example, the Chromatis solution provides a
comprehensive package for transporting integrated video, voice and data through a regional
distribution network. Although the architecture is optimized for digital-video
transmission (including video-on-demand), it can also deliver high-speed Internet access,
as well as Internet-protocol and switched-voice services.
In other words, the same fiber ring -- as envisioned --
becomes a huge highway: a 40-gigabit pipeline for whatever type of bandwidth is needed.
Chromatis -- which has about 80 engineers in Israel writing code, plus about 20 marketing
and technical personnel, mostly in suburban Washington, D.C. -- plans to launch a field
test of the service on a cable system that has been signed, but not yet disclosed, by
The company's technology is centered on its patent-pending
"Selective Wave-Division Multiplexing" technology, which, it claims, drops the
price of WDM systems and, at the same time, is optimized for the digital world.
Not surprisingly, Chromatis' presence in Los Angeles this
week does not indicate any preference toward the cable industry. The company is also
trying to sell its approach to telephone companies and especially to competitive
local-exchange carriers that also foresee markets for a big pipeline that can pump video,
data and voice around a metropolitan area.
Redback, which also makes its cable-show debut this week,
supplies equipment to manage traffic on metropolitan-area networks. Its target is
optical-network operators that carry IP material.
Redback paid $4.7 billion last month for Siara, which plans
to make Internet switching equipment, but which is admittedly at least six months away
from rolling out its first product. (In true Internet-era economics, Wall Street heartily
approved of the multibillion-dollar price for a company without a product.)
The underlying concept of Redback's business is strikingly
similar to that of Chromatis. And others are on the way, too.
But these newcomers -- each of which would desperately love
a clustered cable network to show off its integrated capabilities -- are not alone in
their pursuit of the big-pipeline approach.
Cisco Systems recently showcased its latest gateway
product, which integrates conventional circuit-switched networks with ATM and IP networks
for voice traffic. Cisco's further digital-video-data products will join the parade this
Now that's the birth of bandwidth-on-demand.
These systems give cable and telco operators great
flexibility in adapting services to whatever customers want at any time. Ultimately, they
are cheaper than dedicated SONET or video-transport systems. This gives an almost
immediate benefit to cable operators, which are largely unburdened with SONET technology.
Telcos are still amortizing that investment.
Meanwhile, data LECs can plow into the expanded business
with this technology, especially in a digital-subscriber-line environment. DSLAMs
aggregate cooper loops and feed them into the ATM-switching matrix, again providing better
utilization of the bandwidth.
All of these technologies are built on the expectation for
next-generation networks -- which wind up simply being one big integrated network. For the
newly clustered MSOs, it offers immense competitive clout if they can build and manage the
new highway. Theoretically, the existing telco will be lumbered under legacy systems that
don't intertwine or offer the flexibility of this new big pipeline.
The newcomers are promising lower costs because of the
centralized router and equipment. Moreover, the setup allows consolidation of headends
around the metropolitan area -- effectively "unmanning" (a term that makes you
wince as you cross your legs) the redundant sites around town.
The big-pipeline approach hooks into distribution hubs,
which then pump bandwidth down existing neighborhood networks -- presumably hybrid
fiber-coaxial systems -- toward the other emerging component that will be highly visible
at this week's cable show: the home gateway.
It might be a digital set-top box, or, as many competitors
will try to show, it could be some other structure to handle the gushing video, voice and
data pouring through the last few meters at the end of that big digital pipeline.
Now cable operators have to decide how to handle this
integrated gusher that they've been promising for nearly a decade -- and also how to deal
with this new breed of vendors that are eager to supply it.
I-Way Patrol columnist Gary Arlen expects to continue
playing in high-speed traffic well into the next siècle.