Press Release

March 2001 - Catching the Seventh Wave

Startup Wave7 Optics dives into fiber-to-the-home
BY ROGER BROWN, EDITORIAL DIRECTOR

As last summer’s blockbuster film, “The Perfect Storm,” taught us, big waves aren’t necessarily things you want to tangle with. But get on the good side of a killer wave, and you might sweep past your competition and get home first.
Executives at Atlanta-based startup Wave7 Optics (named after the nautical axiom that says the seventh wave is always the largest) hope that their company can ride the crest of this wave all the way to market success.

For at least the past decade, we’ve been told that the next “big thing” in the optics arena will be the advent of fiber-to-the-home. But over that time, it’s been held at bay by high equipment costs, craft issues related to the last-mile and networks that were able to leverage existing copper and still provide enough bandwidth.
Cable operators, in particular, have seen little need to drive fiber all the way to the home (though they are pushing toward 100-home nodes) because of hybrid fiber/coax networks that have largely satisfied the need to deliver broadcast video and high-speed data downstream.


But breakthroughs that have occurred in the optical world, combined with the advent of new, bandwidth-consuming applications like Napster, have caused a few old cable TV pros to dive into the FTTH pool head-first.
Industry engineering icon Jim Farmer has teamed up with former Antec colleagues Emmanuel Vella, John Kenny and Tom Tighe to found Wave7 Optics, a company that thinks it has found a way to cost-effectively deploy fiber-to-the-home or business, whether or not an operator has already invested in HFC or similar technology.
“Everything appears to be doable,” says Farmer in his typical conservative manner. “We’re using off-the-shelf components. We’re using the latest and greatest, but we’re not counting on any technological breakthroughs (to make this happen),” he adds.
The problem with FTTH architectures of the past were many, according to Vella. There was high initial cost for equipment and installation because it was big and bulky and required large amounts of real estate. They also required high fiber counts, which led to high connectorization and splice costs. There were also issues related to distance (the signal “reach” is dramatically reduced because of splitting and cable losses amounting to about 20 dB).
Typical passive optical networks (PONs), in particular, often omit video transmission (instead focusing on voice and data) because the environment is rough on video over RF carriers.
To compensate, alternative PON architectures have been proposed that place an “active” router-type device in the neighborhood. This helps reduce fiber counts on the downstream side of the active device, but it is possible to have sheaths that include several hundred fibers.
The Wave7 approach, however, includes an active device called a “core” that functions much like today’s “node” in an HFC architecture. Instead of a coaxial output, the core has fiber cables emanating toward the home. But because an optical tap is used at the point of the drop, high fiber count cables aren’t needed.
Using this approach, video is delivered in the 1550 nm window using an externally modulated transmitter and erbium-doped fiber amplifiers to broadcast the video to every home. IP-based voice and data signals are sent in the 1310 nm window, using low-cost digital lasers, and are handled by low-cost transceivers available from several sources, according to Farmer, the company’s chief technical officer.

The core device combines the video with the data signals, and telephone signals receive a higher priority than Internet signals in order to assure quality of service. Upon reaching the tap, the signals are split to each home.
The final piece of the puzzle is the home gateway, which separates the data and video streams and routes them to the proper locations. Enough signal level is provided to accommodate a four-way splitter and long drops without additional amplifiers. If the gateway device is a set-top that requires an RF return path, an optical-to-digital converter grabs those signals, digitizes them, and sends them with the other return data.
The basic data interface provided is a 10/100BaseT, which means it’s not necessary to provide DOCSIS modems at the home, or use a CMTS in the headend.
Because the architecture has symmetrical upstream and downstream bandwidth, network operators aren’t forced into using the relatively small return channel that inhabits the 5 MHz to 42 MHz region. This will become a key selling point as peer-to-peer networking applications like Napster become more popular. It also allows users to host Web servers in their homes, another application that’s growing in popularity, according to Farmer.
“The demand for bandwidth has grown tremendously,” says Tighe. “Many cable system engineers are looking at that return path with dread,” wondering how they’re going to be able to keep up with demand, he adds.
In addition to the architecture, which is patent pending, Wave7 has developed some novel approaches to dynamic bandwidth allocation and element management that are also being considered for patents, according to Farmer.
Background
Wave7 came together in September 2000 with $9 million in seed funding from Lucent Venture Partners and Morgenthaler Ventures, and as of last month, had 48 employees, most of whom are design engineers. The company is expected to grow to about 60 people through the product design phase, which is expected to last until at least the middle of this year, according to Vella, the company’s chief marketing officer.
A second round of financing is expected to fetch upwards of $20 million and should be closed by June, says Tighe, Wave7’s president and CEO. Notably, both Lucent and Morgenthaler are expected to up their antes in that funding drive.
Farmer and company expect to be able to demonstrate the optical portion of the network sometime in March. In May, hardware and software will be combined and, if all goes according to plan, field trials should commence by the end of this year, leading to a product roll-out early in 2002.
The company’s key to success will be whether it can deliver a low-cost approach as promised. Tighe says his approach should be on economic par with HFC gear, based on 100-home nodes, by next year. Considering the system can provide so much more bandwidth than HFC, that would be a powerful selling proposition. “We’re developing a cost-effective fiber-to-the-home or fiber-to-the-business platform that comes in at the same price as HFC, but has 300 times the bandwidth,” he says.
The backbone’s bandwidth is variable, depending on how much redundancy is built into it, but Vella says that up to 4 Gigabits per second can be shared within a 96-home cluster, and that burst rates of 450 Mbps per home can be achieved.
Farmer notes that such an approach can peacefully co-exist with HFC networks, although it’s probably not cost-effective for a cable operator who recently upgraded his plant to immediately go with this approach. “It does lend itself well to major rebuilds and plant extensions, however,” Farmer says. “All you’d have to do is abandon the old coax.”
Farmer admits there are still some issues related to splicing and powering, but believes there are work-arounds. “Craft issues have been a concern, but we think we have some ways to approach it with connectors and mechanical splices. The system will actually be simpler than HFC, because there’s no need to adjust for tilt or use pads.”
On the powering side, traditional network power can be used (and Wave7 has worked out a way to place two active cores in the same location to take advantage of a single power supply) or operators can choose to use home-based units, much like AT&T has chosen to do.
For Wave 7, the next few months will be crucial as it builds its first products. Two or three limited, targeted trials are already being mapped. “We’re taking a cautious approach,” notes Vella.

For more information please contact Will Bryan

<Back>