Next generation SONET/SDH stays the course
Since its emergence several years ago, the so-called “next generation” SONET/SDH market has continued to evolve from something much like legacy SONET/SDH equipment to a piece of equipment now known as the multiservice provisioning platform (MSPP). “ ‘Next generation’ is very broad in the sense that it applies to any new generation of SONET equipment that adds features for data,” explains Jimmy Yu, senior analyst with the Dell’Oro Group (Redwood City, CA). “Originally, an MSPP meant offering different types of client support at the metro edge of the network.” Both terms as they were originally defined may have outlived their usefulness, however; what was deemed next generation SONET/SDH a few years ago is no longer next generation SONET, and the functionality of MSPPs is expanding beyond the metro edge.
The SONET/SDH market today accounts for two-thirds of the optical market and will continue to do so going forward, asserts Jacob Larsen, director of portfolio management and business development in Optical Networking Group of Lucent Technologies (Murray Hill, NJ). “The resiliency-the saving grace-of SONET/SDH has been the protocol’s ability to incorporate other network functionality,” he says, citing the adoption of generic framing procedure (GFP), virtual concatenation (VC), and link capacity adjustment scheme (LCAS) as the key advancements that have enabled SONET/SDH equipment to efficiently handle data, which in many people’s minds today means Ethernet traffic.
But that doesn’t mean multiservice SONET/SDH boxes are predominantly populated with Ethernet ports-at least not yet, anyway. “The MSPP was designed assuming there would be more Ethernet ports than TDM-type ports, but that hasn’t really materialized,” reports Yu. “Actually, a very small percentage of Ethernet ports ever appear on these multiservice boxes.”
Kevin Drury, senior marketing manager at Optical Network Division of Nortel Networks (Ottawa, Ontario) agrees with this assessment: “When you look at the number of ports being deployed in any one of these MSPPs or next generation SONET platforms, the overwhelming majority of them will be DS-1s and DS-3s.” But he cautions that this fact may be a bit misleading. He says it’s important to look not only at the number of tributary cards and number of ports on those cards, but also at the amount of bandwidth they drive on the backend. “Even though the actual Ethernet cards being sold may be a small percentage of the card volume,” he explains, “the entire amortization of the common equipment on that shelf-the OC-48 cards, the common equipment, the shelf processor, the switch matrix-all of that stuff is consumed and being driven by the data cards.”
The metro core boxes may be predominantly populated with DS-1s and DS-3s, but a new class of boxes has emerged to target the enterprise space where Ethernet is prevalent. The mini-MSPP is “a smaller, purpose-built, right-sized device that can sit at the very edge of the transmission network or at a customer location and provide the multiservice capabilities that a true MSPP delivers today,” reports Scott Messenger, product marketing director of Cisco Systems’ Optical Group (San Jose, CA). “The only difference is that the cost points, size, and power consumption are traditionally much smaller.”
Yu estimates that the mini-MSPP market currently accounts for around $20 million per quarter but that revenue should increase as the devices are more widely deployed. “Going forward, I see a transition from the legacy OC-3 and OC-12-type products to these newer platforms,” he says. “The multiservice transition actually occurred at the OC-48 rate, not the lower line rates. The closer and closer you get to the edge of the network, it’s still legacy equipment. When it comes to backhauling DSL, that’s still all legacy SONET gear.”
Adds Larson, “The near-term trend is really penetration of the optical network much closer to the endpoints of the network, to the enterprise for delivery of Ethernet services but also almost, and I underline ‘almost,’ penetration into the residential market.” He believes that SONET/SDH will become the protocol of choice for backhauling fiber to the premises (FTTP). The challenge will be to reduce the cost of SONET/SDH equipment enough to make it a viable option for backhauling broadband access.
In the long term, Larson regards the movement toward optical and data convergence as a growth driver for next-generation SONET/SDH. To achieve this convergence, vendors must marry the flexibility and price points of the data network with the resiliency and reliability of SONET, a task he believes can be achieved, thanks to advancements at the component level. “This is the next battleground,” he contends. “If optical penetration is a small hill in terms of the next challenge, then optical and data convergence is beyond any doubt like scaling Mount Everest.”
Cisco’s Messenger reports seeing another possible long-term growth driver: the need for reconfigurability, particularly in the metro DWDM space. The end game, he says, is to improve service velocity for the customer, and reconfigurable optical add/drop multiplexers (ROADMs) promise to do that. “One way to improve service velocity is to give the customer more flexibility to take that service and map it to any wavelength regardless of the equipment that may be out there, and that is effectively what these ROADMs do. All [the customer] has to do is order one unit, and that unit has the ability to switch to any of 32 channels or wavelengths. When you introduce this service in ring-based architectures, this amount of flexibility increases by a magnitude.”
The big question, then, is how the emergence of ROADMs will impact the future growth of the next generation SONET/SDH market. “Right now, the majority of the long-haul network is DWDM,” observes Yu. “There’s no reason why you can’t just move that a little bit closer to the edge. Interconnecting the major central offices with DWDM in the metro, being more DWDM and less SONET-it’s definitely possible.” However, he cautions that ROADMs still must prove their reliability and viability before they can begin to challenge the MSPP market share. “But it definitely moves the line drawn in the sand between when you deploy SONET and when you deploy DWDM,” he admits.
The SONET/SDH vendors understandably are not convinced. According to Larsen, ROADMs and next generation SONET/SDH boxes tackle different problems: “ROADMs are there to solve the fundamental lack of flexibility in fiber and, conversely, in DWDM. What SONET/SDH allows you to do is groom the traffic, multiplex the traffic-in industry lingo, we call it traffic engineering. Grooming and multiplexing traffic gives you very, very large benefits in optimizing your network. At the most fundamental layer, I see those two things being totally noncompetitive,” he says.
Nortel’s Drury believes that simple economics will force ROADMs and MSPPs into a complementary relationship. Next-generation SONET/SDH lasers in the OC-48 1310-nm range are less expensive than 1550-nm DWDM-type wavelengths, he notes. “Until the whole wavelength industry comes down in cost, I don’t think that ROADMs as a platform will make any significant inroads as a substitution technology for your next generation SONET/SDH,” he says.