Israeli Vendor Pitches Ops On Tool to Xtend Bandwidth

Figuring that high-definition TV and blooming video-on-demand products will eventually burst the seams of a standard 750-megahertz cable plant, an Israeli outfit is hitting U.S. shores with technology that can boost hybrid fiber-coaxial bandwidth to a whopping 3 gigahertz for as little as $100 per subscriber.

The question is: Is that too much, too soon?

Xtend Networks Ltd. kicked off its marketing efforts last month by announcing that an cable operator in Israel would deploy its technology. The vendor showed off its wares last week at Society of Cable and Telecommunications Engineers's Cable-Tec Expo in San Antonio.

Xtend's technology focuses on the physical-layer connections in the last-mile coaxial cable, essentially creating an overlay bandwidth path above the existing 750- or 860-MHz network.

It does so by latching on new components at key areas of the existing plant — an XHUB interface set at the fiber node to upconvert signals to the higher range; an Xtendifer bypass device hooked to each laser amplifier, typically set at every 100 to 300 yards of coax; new higher-capacity passive splitters; and, at the consumer premise, a downconverter device that translates the higher-frequency signals into the digital set-top or cable modem.


That may sound like a lot of extra parts to add, but co-founder and CEO Hillel Weinstein claims it can be done for no more than $100 per subscriber, including components and installation. In a recent beta test in Israel, one neighborhood node was installed and brought online in about six hours, Weinstein noted.

"We are only operating at the connector level," he said. "We don't dig, we don't split, we don't cut. We leave the cables where they are, we leave the components where they are."

So far, Xtend's tests have found that its technology can inadvertently improve lower-frequency transmissions as well.

"We very often find that when we replace those splitters, not only do we transmit and receive the information over our extended band, but also the conventional CATV up to 860 improves, because we have replaced certain components with better-performance components," Weinstein said.

The higher-frequency overlay also doesn't require additional power supplies, drawing instead from existing network resources.

"From a topography point of view, it uses existing power sources," Weinstein said. "From a component point of view, we operate at lower power levels with our high-frequency devices than the 500 to 860 devices."


Xtend plans to enter the U.S. market through smaller operators and original-equipment manufacturer agreements with the larger MSOs.

"Eventually we are also entertaining the thought of working with the hardware OEM community in the United States," Weinstein noted. "We don't overrule the possibility that we may license some of the technology to the big manufacturers."

Xtend will start manufacturing its devices in August. The company is negotiating with some high-volume manufacturers to start mass production by early 2003, Weinstein said.

But are cable operators really looking for 3 GHz of bandwidth? Not really, Weinstein admitted.

"Right now they tell us that if we give them 1.5 to 2 [GHz] that is good enough," Weinstein said. "But we are fairly convinced that as video-on-demand spreads, as interactive television spreads and, last but not least, as HDTV begins to make an impact, they will need the 3 gigahertz."

Nothing prevents the technology from pumping 1 GHz on a symmetrical basis, and that could open cable networks up for business applications, Weinstein noted.

But once again, the pressing question is, do cable operators really want 3 GHz of bandwidth?

"And the answer would be, 'no,' and especially in that format," said Michael Harris, president of cable analyst firm Kinetic Strategies Inc.

"Right now, there is no downstream bandwidth constraint," he noted. "As VOD starts to get penetrated over the next two to three years, we'll have a better sense of whether they need more capacity.

"But most of the models that the residential operators have at least for residential services shows that the 750 that they have is plenty. And the cable industry is incrementalist. If they need more, they will buy it then."