cover story
So many PONs, so little market?
Many forecasters think fiber-to-the-home and -business won't be a significant market for another three to five years. Many passive-optical-network vendors say they're wrong-and claim they have the customer interest to prove it.By Stephen Hardy
Someday, nearly every home and business will send and receive voice, data, and video over a fiber-optic cable that reaches right to the front door. Of course, a cartoon character once sang that someday her prince would come. Based on the costs traditionally associated with fiber in the access space-and the amount of investment the biggest players in that space, the regional Bell operating companies (RBOCs), have already sunk into their existing copper-based infrastructures-industry observers have long considered fiber-to-the-home and -business just as much a fantasy as "Snow White."
However, the sudden spate of optical-networking vendors focused on fiber in the access has challenged this notion.
Passive optical networks (PONs) can economically bring high-capacity fiber links directly to homes and businesses, say such companies as Quantum Bridge, Terawave Communications, AllOptic, and Optical Solutions, not to mention more mainline vendors such as Fujitsu and NEC (through its NEC eLuminant arm).
Based on prevailing wisdom and previous market forecasts, one might conclude that these companies are running ahead of the market to stake out positions from which they can sell their wares in a few years when carrier requirements finally catch up with their technology. However, it has begun to appear that, much like in other areas of optical networking, carriers are moving to fiber-in-the-local loop faster than analysts had expected. Cheer up, Snow White; your prince may be on the way.
Active on passive
While vendors offer different flavors of PON, these architectures share basic components. A unit in the central office or headend, frequently called an "optical line terminal" (OLT), initiates service toward the end user. A passive optical splitter between the OLT and the end user splits this stream into multiple transmissions, which are sent to terminating devices either at the subscriber (in the case of fiber-to-the-home or -business) or at a node somewhere in the local loop (for fiber-to-the-curb and similar architectures). The term inating devices are frequently called "optical network units" (ONUs). PONs provide service both upstream and downstream, making them capable of supporting multiple, interactive services.The RBOCs have been pioneers in this area, despite their more visible efforts in DSL architectures. BellSouth, SBC Communications, and US West joined with carriers and vendors from around the world in the Full Service Access Network (FSAN) initiative, which resulted in the definition of technical specifications for an ATM-based PON. These specifications have been accepted by the International Telecommunication Union-Telecommunications (as ITU-T G.983.1), the ATM Forum, and the European Telecommunications Standards Institute. BellSouth (as well as such international brethren as British Telecom, France Telecom, and NTT) has launched FSAN fiber-to-the-home trials, using equipment from Lucent Technologies and Oki Electronics. (For more on FSAN, see Lightwave, September 1999, page 60.)
The FSAN confab hoped to establish a standardized method of building PONs, both to enable interoperability between networks and promote multiple sources of equipment. RBOCs generally have been steadfast in at least paying lip service to the necessity of PON compliance. For the most part, vendors have taken heed, even among the startups. Terawave, for example, has touted FSAN compliance in its offering, even though the company's system provides features not covered in the standard. (Additions to the FSAN standard are forthcoming, say Terawave sources; the company will present its technology for possible inclusion in these new specifications.) Optical Solutions, meanwhile, plans to bring its equipment into compliance with FSAN by next year to promote interoperability, according to the firm's chief technical officer, Bob Lund.
 |
Still, FSAN certification isn't as much of a requirement as one might think, Lund says. "The people who are deploying fiber-to-the-home, to varying degrees, are interested in FSAN, but they are not letting that stop them from deploying products. The people that are farther away seem to have more of an interest in FSAN," he offers. "The RBOCs, to varying degrees, have certainly articulated a desire to deploy FSAN-compliant equipment. I don't know when push comes to shove if it's not there and they have the need whether that would stop them or not."
Bob Larribeau, director of access networks at market research and analysis firm RHK Inc. (South San Francisco), agrees that an open mind would well serve any carrier looking at fiber-in-the-local loop. "I think that in this area that the technology is fairly new and that carriers are pretty much likely to be committing to a single vendor," he advises. "And they should really be considering the services they offer, and the manageability capabilities of the particular solution should have a very high weight." That means carriers shouldn't overlook a system that meets their requirements just because it doesn't comply with FSAN.
Where's the need?
Of course, before carriers start to ponder whether FSAN compliance is a requirement, they have to decide that a PON architecture-or even fiber itself-makes sense for their applications. In addition to the DSL alternatives now touted by the RBOCs and some competitive local-exchange carriers (CLECs), the hybrid fiber/ coaxial (HFC) architectures popular among cable-TV companies and wireless systems also promise to meet expanding bandwidth requirements in the local loop. A significant amount of revenue hangs on such decisions; according to ElectroniCast's new "Broadband-to-the-Home Services, Equip ment, and Components Market Forecast," carriers reaped $48.1 billion in 1999 from providing services to residential users in North America. The figure is expected to reach $174.8 billion in 2009.Larribeau believes the future looks bright for fiber. "I think there's a huge market for fiber in the access," he says. "The demands for higher-speed services are huge." Other analysts agree. Communications Industry Researchers has predicted that the market for fiber-in-the-local loop will reach $961.1 million this year, with fiber-to-the-building leading the way. The market should grow significantly between 2002 and 2004, according to the firm; the number of homes connected to fiber is expected to grow to more than 180,000 by 2004, for example. Meanwhile, ElectroniCast predicts in its new report that AT&T will begin "significant new fiber-to-the-home trials" this year and next in North America. As the accompanying Table shows, ElectroniCast sees fiber having an increasingly significant impact in the access space. But just because fiber will become more plentiful doesn't mean that PONs will be the prevailing architecture. "I think that PON is one alternative, competing against other fiber technologies and competing against wireless technologies," Larribeau asserts. "I don't think it's really stabilized yet in terms of what is the best alternative." ElectroniCast sees a future in passive star architectures as the most likely fiber scheme to approach cost parity with copper. That said, even some of the more active carriers haven't made up their minds. For example, at the same time it is running its FSAN trials, BellSouth is experimenting with a cable-TV-style "fiber-deep" architecture based on equipment from Marconi. Larribeau believes that PONs will face significant competition from architectures based on digital loop carriers (DLCs), particularly in RBOC networks and those of other incumbents. "[DLC is] widely deployed in some of the networks. In BellSouth, approximately half their access lines are delivered over DLC; for US West, it's somewhere around a third," Larribeau points out. "I think that's got quite a bit more flexibility than a PON system, although it costs a lot more."
According to Lund, the competition in the fiber-to-the-home market where Optical Solutions specializes comes initially from HFC, particularly if the carrier wants to provide more than Internet access. "If they want to do video, that drives them very quickly to at least look at HFC. But when we talk to people, they would rather do fiber," he says. "What we see most besides HFC is fiber-to-the-curb with VDSL."
 |
| Architectures based on digital loop carriers represent a significant competitor for passive-optical-network approaches-and a lucrative market for communications companies. The market for this equipment was $2.8 billion in 1999, according to RHK. |
|---|
Lund reports that other carriers start their discussions with bit rates, particularly if they have a data focus. "What the discussion usually evolves into is that there's a critical mass from a service perspective. What we are finding is that the people who actually deploy our system-and, by extension, would deploy any fiber-to-the-home system-end up being interested in, does it support telephone service, does it support video delivery, and does it support data service," he explains. "And then in the context of data, of course, they're interested in what kind of speeds can you deliver." This focus on the efficient delivery of multiple services represents a relatively recent phenomenon, at least as far as boosting the use of PONs, Lund says. "Six months ago, people were a lot more interested in [high speeds]; if you don't have 100 Mbits/sec or you don't have 1 Gbit/sec-you can kind of pick the number, depending on who you talk to-they kind of lost interest. As we've moved into these past six months, people have become much more interested in, will it deliver all the services," he reports. "Yes, data is important, but the fact of the matter is that the competition is some flavor of DSL or hybrid fiber/coaxial. And residential customers are taking comparable speeds, on the order of megabits per second, not hundreds of megabits per second. So there's a threshold for data; once you pass that, then it's, do you support these other services in addition to data."
Delivering these services in a cost-effective manner remains a priority. This factor becomes even more crucial when talking about residential applications. "I think the key differentiator for home markets is the customer equipment and the ability to offer that for a very low price," asserts Larribeau. "You have to get it down to the $200 or $300 price range so it will be competitive with a cable set-top box before it's going to penetrate that market."
PON vendors have attacked this issue forcefully. For example, Optical Solutions designed its ONUs to be compatible with standard home and computer wiring schemes. That means consumers don't necessarily have to pay extra for an interface box. "What we have seen so far is that they are bundling the cost of the network termination device that mounts on the outside of the home into their capital equipment expenditures for doing the build," Lund says of his company's carrier customers. "At least insofar as they have shared business plans with us, it does not appear that they are jacking up the monthly bill to try to recoup that."
The bottom line is that the market is growing much faster than analysts have predicted, Lund claims. He asserts that his company expects to hit the deployment numbers forecasted by some researchers for 2004 by next year. Growth has surpassed Optical Solutions' own expectations; the company is already five times ahead of where it expected to be at the end of this year. The problem with fiber-to-the-home forecasts, says Lund, is that analysts focus too much on the RBOCs and not enough on the CLECs, utilities, and other emerging communications services providers who are looking at beating the incumbents to lucrative niches via advanced optical technology.
Laying the cable
Integrated Broadband Network (IBN), a nationwide carrier based in Indianapolis represents one such aggressive CLEC. The company has targeted "greenfield" residential areas as markets for high-bandwidth, fiber-based architectures for the delivery of voice, data, Internet access, and cable TV.IBN started with a market target a little more than three years ago but not an infrastructure, according to company officials. "The real bandwidth shortage was in that last mile going to the home," says Kirk Smith, IBN's vice president. "For a lot of new emerging technologies at that time-security, Internet, video, and things of that nature-we began to think why not try to get fiber into the community, or at least to the curb. After various incarnations of the concept, we finally came up with the idea that fiber was the best way to go-not only from a marketing standpoint, but from a long-term operational standpoint."
However, the executives weren't entirely sure that a fiber architecture would fit their requirements. "Obviously, there's not an issue with fiber having enough bandwidth," explains Sheldon Phelps, president and CEO of IBN. The key was having systems that could accommodate the transmission of multiple services over the same fiber. "Initially, we thought we would only be able to provide data, and that was relatively easy. Then we became aware of different ways that we could do cable for television over that fiber infrastructure," he says.
Eventually, IBN settled on a PON architecture using equipment from Optical Solutions that would provide voice, cable TV, and data over the same infrastructure, while providing an upgrade path to higher speeds when that becomes necessary. "From a cost standpoint, it was very advantageous," says Phelps of the PON architecture. "We didn't have to worry about active nodes in a community of any specific size. We're currently working on projects with homes that are in the 1,300- to 1,900-home range. And we can get more than that amount of density and have it remain a passive optical network. It also allows for a great deal of savings on ongoing maintenance."
Phelps says that the ability to move all three services over the same network, as opposed to operating multiple networks in the same area, particularly appealed to IBN. "We liked the life of the fiber and the maintainability of the fiber as opposed to copper or coaxial," he adds.
As mentioned previously, PONs are flexible enough to accommodate a variety of architectures, from fiber-to-the-cabinet all the way to fiber-to-the-business or -home. Not surprisingly, given his company's orientation, Lund asserts that companies like IBN are best served laying fiber straight to the end customer, as opposed to marching it slowly through the loop as user bandwidth demands increase. The beauty of a PON, he says, is that you can increase speeds merely by upgrading the end-point equipment while leaving the basic infrastructure intact. Creeping fiber through the local loop requires continual infrastructure swap-outs and higher costs, he asserts.
Of course, that may mean a carrier will have more capacity than it knows what to do with at first. The IBN executives appear unconcerned about this possibility. "Even though we're able to provide today, using Optical Solutions' equipment, phenomenal bandwidth to the home, are we going to use that yet? Maybe not this morning," explains Smith. "However, with voice over IP just a couple of years down the road from being widely used, as cable and the computer become fully merged, we're basically laying an infrastructure that we won't have to touch, as these things continue to become pervasive in our market."
Passive aggressive
Despite the claims of success from companies such as Optical Solutions, debate continues over the short-term future of fiber in the access space in general and PONs in particular.Several analysts-and as many vendors-see business customers as the drivers for whatever success PONs will enjoy. "It seems to me like the PON can maybe be a little bit more cost-effective and may be more appropriate for CLECs who are trying to move fiber out into some business areas, or an ILEC [incumbent local-exchange carrier] who has some specific distribution issues that they want to address," says Larribeau in shaping up the technology's competitive position. Therefore, he doesn't see PONs sweeping the field just yet. "I haven't seen any signs that it has really caught on in any huge way," he continues. "So I think that it's still at an early stage, and it remains to be seen whether it's going to continue to be used for specialized applications or whether it's going to catch on for broader deployment."
Technologies continue to improve across the board, says Phelps, and carriers should keep an open mind when trying to match a product line or a medium to their requirements. One factor to consider is that how easily an architecture interfaces with a customer is as important as how well it integrates with a carrier's network or business plan. "They don't want to buy a special telephone or a special TV or anything separate," says Phelps of consumers. "They want to plug their same old telephone into the wall."
How well PONs allow end users to increase their communications capabilities transparently will undoubtedly prove a deciding factor in the technology's future success.