By ROBERT PEASE
Passive-optical-network (PON) technology offers many advantages for bridging the bandwidth bottleneck to deliver true broadband services to business customers over a single platform. But questions remain regarding the technology's maturity and whether all of its various incarnations-Ethernet PON (EPON), ATM PON (APON), and Gigabit PON (GPON)-will see significant deployment.
Generally speaking, PONs are point-to-multipoint local access networks capable of providing a large amount of bandwidth over fiber using passive components. Active electronics are only required at the endpoints of the network. The network is capable of carrying voice, video, and data signals from endpoint to endpoint without a need for optical-electrical-optical conversion.
APON is based on ATM and international standards and has been under development longer than other varieties of PON. Because ATM is a proven, mature technology, carriers interested in PON are currently leaning toward ATM PON to get the first "nods" for potential deployment. The problem is that APON was not optimized for Ethernet-based data services, leaving the door open to EPON.
"EPON systems will play an increasing role as the technology matures," says Gaylord Hart, vice president of marketing and business development at the NEC Eluminant Technologies (Chantilly, VA) broadband access division. "Initial deployments will be largely focused on delivering data services. Ethernet-delivered data services are rapidly gaining momentum in the data services space, largely because of the growing bandwidth requirements of homes and businesses."
Over time, says Hart, Ethernet quality-of-service (QoS) issues will be resolved and EPON standards will be created. At that time, EPON will see much wider deployment for a full range of services, particularly voice and data.
GPON technology was introduced by FlexLight Networks (Kennesaw, GA) as an alternative to APON and EPON. The company claims that although EPON is a step in the right direction, it still lacks two necessary ingredients. First, operations, administration, maintenance, and provisioning (OAM&P), well defined in APON, is poorly defined for EPON. Second, the QoS functionality needed for data networks cannot be accomplished without adding cost and complexity to the network.
"Gigabit PON allows service providers to build networks that are over 90% efficient in bandwidth use," says Gary Lee, vice president of marketing and sales at FlexLight. "This higher level of efficiency, combined with the fact that there is no need for ATM switches at the metro edge-nor any voice-over-IP (VoIP) gear at any end-create a strong case for GPON."
As PON manufacturers battle for customers in a culture of "no technology for technology's sake," the timeline for PON to make a splash in the optical-networking industry has been altered as carriers tighten their purse strings. "The economic downturn has slowed the market," says Peter Jew, vice president of marketing at Optical
Solutions (Minneapolis). "However, there are also positive signs that some operators are expediting their decision for PON because of the much lower construction and material. EPON can be very attractive if one only intends to support data. Operators will continue to deploy APON to support legacy voice and video applications in addition to broadband data."
Setting the stage for PON
PON appeals to the carrier that wants to offer a simple, reliable, low-cost platform offering multiple services to business customers. It provides significantly greater bandwidth than DSL and hybrid fiber/coax (HFC). The fiber offers futureproof infrastructure as bandwidth demand increases. PON can also support legacy and future voice, data, and video applications. But there are still some hurdles to clear.
"The current regulatory environment, particularly the unbundling issue, is not conducive for the RBOCs to invest in broadband access technologies such as PON or NGDLC [next-generation digital loop carrier]," says Jew. "The time will come when the RBOCs will inevitably have to play catch-up and deploy PON to compete with the cable companies."
Another issue is the maturity of the technology and its acceptance by carrier customers. Service providers are reluctant to deploy technologies that require major network overhauls and replacement of legacy infrastructure. Although one of PON's major attributes is its cost-effectiveness, there is still the cost of replacing copper with fiber when it comes to the access space.
Andrew McCormick, director of optical networking at Optical Strategies, a Boston-based telecommunications analyst firm, says fiber cable must be driven further into the network and fiber-to-the-x must become a more widespread phenomenon. "PON will, in a limited manner, be more of a following technology for breaking down the access bottleneck," says McCormick. "The answer to the access bottleneck must be copper-based. Technologies that can provide higher bandwidth at longer distances-10-12 km-will compete with PON. I'm not sure PON will come out on top."
McCormick believes that despite ongoing trials and evaluations by service providers, it will be a long time before PON is deployed in any quantity. By then, the economic crisis will likely be over and multiple service operators (MSOs) will probably begin using PON to some extent. "I think it will continue to be most popular with rural local-exchange carriers [LECs] and cable providers that are upgrading cable plant to serve small towns and counties and expand their service areas," he says. "Optical Solutions has carved a nice niche here." Optical Solutions currently has over 32 commercial deployments of PON. NEC Eluminant's Hart agrees that PONs will succeed best for business applications where service demand already exists and fiber is already in place.