Skip to main content

TeraBeam, Lucent Extend Bandwidth Limits

Old saws about the cost barriers to providing high-bandwidth access to business users may soon be outmoded by new developments in over-the-air and wireline-based optical systems that go beyond the capacity limitations of fixed-wireless and copper networks.

In one important sign of what's to come, TeraBeam Corp., a Seattle-based developer of free-space optical technology, has teamed up with Lucent Technologies to expand a point-to-multipoint system that TeraBeam's service-provider subsidiary is already using to test 1-gigabit-per-second connectivity with a Seattle customer.

If the system lives up to its creators' claims, Tera-Beam will be able to expand quickly to several cities, offering a low-cost Internet protocol- and Ethernet-based gigabit service that operates reliably even under the most foul weather conditions.

Separately, Comcast Corp. and two other unidentified MSOs have either begun or soon will begin testing a fiber-based optical system supplied by North Andover, Mass.-based Quantum Bridge Communications Inc. That vendor claims that it can enable the extension of high-bandwidth service to enterprises at a fraction of traditional fiber costs.

As is the case with the IP-based TeraBeam system, one of the advantages Quantum Bridge claims is the ability of the tested system to make optimal use of bandwidth by operating in native protocols, rather than having to translate everything to traditional telecommunications time-division-multiplexed modalities.

Convincing companies that they can rely on mission-critical communications via a light beam coming through an office window could prove to be a major challenge, compared to fiber-based or even fixed-wireless solutions.

But TeraBeam and its new backer, Lucent, believe they have a solution that classifies as "disruptive," in the phrasing of technology guru George Gilder.

In fact, Lucent chairman and CEO Rich McGinn said his company, after considerable investment in point-to-multipoint (PMP) microwave and millimeter-wave wireless technology, has concluded that LMDS (local multipoint distribution service) and similar systems will have limited usefulness in the marketplace.

"We haven't seen a lot of industry success with PMP technology so far," McGinn said. "With the speed of provisioning and other advantages offered by the free-space optical system, we believe it is far superior to what we're seeing in fixed-wireless systems today."

Lucent and TeraBeam have teamed up to create Tera-Beam Internet Systems, a technology-development company in which Lucent will hold a 30 percent stake and TeraBeam the remainder.

TeraBeam Internet Systems will supply service provider TeraBeam Networks and, eventually, other service providers with a proprietary means of delivering 1-gbps pipes on a PMP basis from network-hub locations. Each of four sectorized transmitters will be positioned at each hub, capable of reaching dozens of users at such speeds, officials said.

"Our first generation of equipment supports up to 250 gbps per sector, which, with four sectors to cover the full 360-degree cell, gives us 1-terabit capacity per cell," TeraBeam CEO Dan Hesse said.


Hesse estimated that an entire market of high-bandwidth users could be served by 12 to 20 such hubs, depending on the size of the city.

Lucent's affiliation with TeraBeam builds on its move late last year into multiple-wavelength point-to-point (PP) over-the-air optical systems, when it introduced its "OpticAir" product line.

Lucent brings $400 million worth of OpticAir technology to the deal, providing a means of interconnecting TeraBeam PMP hubs, as well as delivering PP services to end-users from those hub locations, McGinn said.

"We at Lucent believe the TeraBeam PMP technology has the potential to literally change the landscape in local-access services to the commercial market," he added.

TeraBeam's PMP technology relies on holographic lensing techniques to shape wide light beams to match the optical-reception characteristics of each end-user's premises equipment, Hesse said.

Unlike the very narrow beams that are used in dense-wavelength-division multiplexing, the TeraBeam output uses wide beams at the same wavelength that don't have to be separated from each other because the special shaping of each beam protects against interference.

In fact adjacent transmitters can be added to augment capacity without causing the type of interference that prevents colocation in fixed-RF-wireless systems, Hesse noted. "Later generations will use WDM, but we're well positioned to meet capacity needs now," he said.

TeraBeam Networks plans to launch commercially in Seattle and possibly other cities by year's end, and to move to 100 cities worldwide within four years, Hesse added.

The service company will be the exclusive customer for the over-the-air system developed by TeraBeam Internet Systems for an undetermined period of time, to ensure that it has all of the gear it needs for its aggressive expansion plans, Hesse said.

"Once we are sure our system needs will be met at TeraBeam Networks, we could allow the equipment to be sold to other service providers," he added.

Hesse said TeraBeam Networks intends to work with various service providers and resellers in local markets, offering a means by which entities such as cable competitive local-exchange carriers could make use of the system without actually owning it.

Hesse declined to discuss specifics of how the proprietary technology works. But he and McGinn made it clear that they see provisioning of bandwidth in 1-gbps slices as cost-competitive with wireless and terrestrial-access systems that deliver far less bandwidth per user.


"There's no spectrum to license, no indoor cable risers to connect with because the signal goes right through the window, and it takes only a couple of hours to hook up a customer," Hesse said.

He declined to say how much TeraBeam will charge for the service, but the customer-premises equipment would be supplied by the service provider as part of the service fee, he said.

By conducting tests in Seattle, he added, the company is demonstrating that the system can work under virtually any weather conditions.

But if Quantum Bridge has its way, it may be too early to dismiss the viability of terrestrial fiber optics as a cost-effective means to expand the bandwidth of fixed wireless and copper.

Although Quantum Bridge doesn't claim to be cost-competitive with fixed wireless, it offers the robust connectivity of fiber at a far lower price than fiber connections have traditionally cost, vice president of marketing Jeff Gwynne said.

The company has worked with Comcast at an undisclosed location since the start of the year to field-test its system, said Gwynne. It will be in a position to begin shipping units for commercial rollouts in June, he added.

Quantum Bridge does not have any commitments for commercial applications yet, but expects the current round of testing to produce such orders, he added.

"Comcast sees an opportunity to leverage fiber in its networks to serve a broader area than it can reach with coax," Gwynne said. "Our system allows it to do that at a fraction of the costs of using traditional means to extend fiber to end-users."

Comcast did not respond to requests for comment about its use of the Quantum Bridge system. But vice president of strategic planning Steve Craddock has made clear that the business market is now a key target for the MSO's use of new packet-voice and data services.

Addressing a session at the recent Voice on the Net conference in San Jose, Calif., Craddock said: "If I'm going to sell to business, I need to be able to say that, using cable-modem technology, I can completely replace their current systems with throughput that's much better than what they're getting today with T-1 lines and manage 56 lines at a fraction of the price."


Quantum Bridge claims to be able to meet this requirement by providing cable and other service providers with a way to use already installed fiber, much of which is held in reserve in metro and distribution conduits, to reach a large number of customers.

Instead of provisioning a single pair of fibers from the headend or central office all the way to the premises, operators can put a Quantum Bridge passive optical splitter at the termination point of an existing fiber pair and extend individual fibers from there to many as 32 locations.

"With this technology, the cost of reaching end-users with fiber falls drastically because you're only having to extend fiber another 1,000 or 2,000 feet from the end points, rather than running dedicated fiber all the way from the headend," Gwynne said.

That translates to a cost for the fiber construction of about $3,000 to $6,000 per customer, versus the $150,000 or so it would take to connect over a five-mile fiber run, he noted.

"The key to being able to do that is what we call dynamic wavelength slicing,'" Gwynne said.

DWS-using Quantum Bridge's optical-access switch and special CPE-relies on the synchronization of a 622-megabit-per-second optical signal so that terminals at each end-user site take out only the bandwidth slice dedicated to that site, Gwynne explained.

This way, a signal transmitted in IP or ATM (asynchronous transfer mode) format is delivered to the customer as a lightwave signal right into the ATM or router port, avoiding expensive conversion of the entire 622-mbps lightwave signal to RF for demultiplexing the dedicated signal.

"Fully loaded, the system will allow network operators to sell a 10-mbps service to each user at the price businesses now pay for DSL [digital subscriber line]," Gwynne said. "You can charge $500 to $1,000 per month and get payback on the costs of connecting the customer in one year or less."

By using "course" WDM at a rate of seven wavelengths per fiber, the Quantum Bridge system also supports delivery of additional services, Gwynne noted.

For example, six different wavelengths that operate in the 1550-nanometer wavelength window could each deliver very high-bandwidth streams to single dedicated end points, while allowing a wavelength at 1310 nm to deliver optically "sliced" signals to multiple users.

This allows the operator to serve end-users with bandwidth needs as high as OC-48 (2.5 gbps). The operator could also sell the bandwidth streams at wholesale to other service providers via additional fiber extensions from the termination of the fiber pair, while providing retail 10-mbps services to multiple end-users via the PON-delivered (passive optical network) wavelength stream.

Developments at Tera-Beam and Quantum Bridge clearly point to a need to rethink approaches to serving the business community.

If these technologies pan out, the cable industry could find it has an extremely cost-effective opportunity to extend its high-speed, high-quality packet-data system into the lucrative commercial market.