AFM the antidote to impending ‘fiber crisis,’ say vendors

Th 0701lw Glimmerglass

With the build-out of the access network, carriers are pushing more fiber ever closer to end users, resulting in central offices (COs) with what one vendor calls tangled “rats’ nests” of patch cords. Several companies have begun touting automated fiber management (AFM) technology as the antidote to congested COs. However, AFM will not see widespread deployment unless it can hit price points carriers can live with, and that appears to be around $100 per port.

There are four principal players in the AFM space, and though each employs proprietary technology, they all purport to be essentially the same thing: a replacement for manual patch panels and fiber distribution frames (FDFs).

AFM equipment is described as more intelligent than alternatives, with some sort of automated fabric or switch that enables a connection between composite fiber A and composite fiber B. In the case of Calient Networks (www.calient.net) and Glimmerglass (www.glimmerglass.com), the switch fabric is microelectromechanical systems (MEMS)-based, while Polatis (www.polatis.com) uses a piezoelectric beam-steering technology to connect fibers. Israeli-based FiberZone Networks’ (www.fiberzone-networks.com) calls its technology Dynamic Optical Coupling, and it uses high-precision motors to maneuver fibers into position.Th 0701lw Glimmerglass

AFM manufacturers are developing smaller, less expensive devices to facilitate widespread deployment. Glimmerglass’s recently announced fourth-generation Intelligent Optical Switch family, for example, supports up to 190 fibers in a 2-U package (shown here with a compact disc) and 380 fibers in a 4-U package.

Existing patch panels and FDFs are inexpensive, passive devices, and therein lies the challenge for AFM equipment: It must be almost transparent from an optical performance standpoint. Both insertion loss and return loss must be low because no carrier is going to recalculate its optical power budget to accommodate a new patch panel.Th 0701lw Diamondwave

Thanks to what Calient Networks calls “significant design and manufacturing breakthroughs,” its DiamondWave FiberConnect platform hits price points that make its deployment in the access network palatable to carriers with extensive FTTX architectures.

Furthermore, AFM equipment must be low cost. According to Glenn Wellbrock, director of network technology development at Verizon Business (www.verizonbusiness.com), Verizon would like to see an AFM system priced around $100 per port. He says one of the biggest issues Verizon faces today is the cost of the bulkhead connector, which is the plastic device that plugs into the back of the patch panel and has a connector on one side and a patch cord on the other. Using standard components, that device is around $100.

Carriers may be willing to pay a small premium for automation, but they won’t pay much more than what they are already paying for patch panels and FDFs.

Calient Networks has begun marketing its products as analogous to DACs in the copper world. Instead of using the term “automated fiber management,” Calient markets its products as fiber-optic cross-connection systems (FOCS), which is really resonating with carriers, says Jim Diestel, Calient’s director of product marketing. “The engineers remember the days when DACs were brought in to ease migration, to mechanize trouble reporting, and to offer some degree of flexibility in the provisioning of the copper plant,” he says.

The carriers’ biggest management problem with copper was deceptively simple: too many patch cords. Wellbrock cites one particular patch panel on 7th Street in New York City as an example. The patch panel itself is two blocks long and at least two stories high-three stories in some places, he says. “No one wants to touch it,” he admits, “because there is active traffic on it. No one wants to cause an outage, so rather than try to recover the old fiber or copper, you just leave it in there. You abandon it, and pretty soon, it becomes unmanageable.”

The list of potential applications for AFM equipment is as diverse as the vendors that manufacture the equipment. Michael McLaughlin, vice president of corporate development at Glimmerglass, narrows the list to two key areas: 1) where frequent changes or 2) very fast changes must be made to the fiber connection.

Applications predominant in the first category include network monitoring, fault isolation, and trouble resolution. Most carriers do not want to have an entirely unmanned CO-though in remote locations this is possible-but AFM does streamline the process enough to enable carriers to meet future service requirements with their current technical staff. “This means bringing more of the job to your experts than experts to the job,” says Calient’s Diestel. “You want more remote access. You want to be able to configure networks from the NOC [network operations center]. You want to be able to diagnose problems from the NOC. You want to dispatch only when you need to, and then dispatch to precise locations.”

When changes to the network configuration are required, however frequently or infrequently, they must be made in milliseconds, which is virtually impossible for a technician to accomplish. In this case, says McLaughlin, carriers use AFM equipment to protect their networks against unplanned outages. In the event of a card failure or fiber cut, for example, the AFM system instantaneously switches traffic to a redundant path, enabling carriers to maintain high quality of service.

With the deployment of more fiber in the access network, ATM vendors are touting their products as the antidote to the looming “fiber crisis,” as Diestel puts it. Carriers such as Verizon and AT&T, which are aggressively pursuing FTTX strategies, “are scratching their heads somewhat about how they are going to manage all that fiber,” he says.

Sandy Roskes, vice president of marketing and business development at FiberZone Networks, reports that carriers deploying FTTX in volume are interested in fiber management to minimize the onerous labor requirements of installing and connecting new subscribers. FiberZone’s AFM-400, for example, would be deployed in locations with a high fiber count, most obviously in the CO. He also sees potential application in the field for carriers deploying a centralized splitter architecture with a number of splitters in a single location. “You could have some hundreds of subscribers supported by or connected in one splitter location,” he says, “and that would be a good candidate for an AFM solution.”

Because of the size of the fiber and the amount of bandwidth it carries, even the smallest bit of dirt or dust can degrade or disrupt an optical signal. As a result, every time a piece of fiber is connected, it first must be cleaned. Deploying an AFM system that automatically switches fiber connections eliminates the need to clean connectors with every change.

AFM in the access network also would facilitate network upgrades more easily, says Diestel. “If I have a physical fiber that goes to an OLT in the central office, and I change that OLT for the next-generation GPON-say, one that supports the 64:1 split ratio instead of the current 32:1 split ratio-I now have to take that fiber off and connect it to another piece of equipment,” Diestel explains. “But if I’ve got an optical switch there, I can then fiber up the second OLT and then just switch over to it. I’ve established it and cleaned it, and then I can quietly and at my leisure disconnect the old system.”

Calient Networks is banking on the access network build-out and has developed several new technologies to help remove some of the cost from its optical switches. When the company first began shipping product in 2002, its optical switches cost between $10,000 and $12,000 per port. Today, the company is selling its switches for less than $300 per port-in volume, $200, says Diestel. He attributes his company’s declining costs to a proprietary feedback circuitry that enables the switch to constantly seek a more optimal connection. Moreover, he says, Calient’s yield has improved from four usable mirror matrices per wafer to 12 matrices.

FiberZone’s Roskes also believes that AFM devices can hit the carriers’ desired price requirements for widespread access deployment, assuming the device contains fundamentally low-cost components. “It must deploy components that are already high volume and therefore enjoy the beneficial cost curves that come with volume,” he says. “And the integration of the components has to be quite straightforward. Products that meet those characteristics can meet the price points required.”

For his part, Glimmerglass’s McLaughlin does not believe AFM systems will become economical in FTTH applications for at least a few years. However, he says, “When you start talking about homes, you’re talking about a very large volume. And when you’re talking about a large volume, you’re suddenly now talking about the ability to get costs much lower because you’re talking about silicon. So it really is a chicken and egg [scenario],” he admits. “Who will blink first? Who will let all of us go to our vendors and say, ‘Okay, let’s give you this huge order. Let’s negotiate on volume.’”

The catalyst, says McLaughlin, may be a couple of large carriers or perhaps just one key carrier like Verizon coming forward and saying, ‘Yes, we’re going to do this.’ When a major carrier like Verizon makes that kind of commitment, vendors up and down the food chain will have more incentive to sharpen their proverbial pencils.

Meghan Fuller is senior editor at Lightwave.

More in PON/Systems