MSPPs follow network evolution
Multiservice provisioning platform (MSPP) technology has evolved such that it now can be integrated into a blade or a chip, leaving one to wonder about the fate of the traditional, standalone device. Moreover, today’s MSPPs include a host of functions, including next-generation SONET/SDH, carrier-class Ethernet, WDM, and reconfigurable optical add/drop multiplexing (ROADM), prompting some to ponder whether the MSPP is becoming perilously similar to the prebubble class of devices rather derisively known as “The God Box.”
The traditional MSPP is a TDM-based switch, which is the same as a SONET switch and therefore satisfies carriers’ need for reliability, explains Eve Griliches, research manager of telecom equipment at IDC (www.idc.com). It handles multiple types of data, including ATM, SONET, and Ethernet, but the underlying structure is TDM-based.
That said, the market is clearly in transition, suffering from what Griliches calls “an identity crisis.” Because MSPP technology has evolved to the point where it fits on a blade or even on a chip, it can be ported to other media, including metro WDM. “So the question is, ‘Does the MSPP turn into an Ethernet aggregation ROADM-like thing? Or do ROADMs incorporate Ethernet aggregation and MSPPs on a chip?’” she muses. For her part, Griliches believes the MSPP market will turn into the metro WDM market, the bulk of which will comprise ROADM-type products.
The folks at ECI Telecom (www.ecitele.com) admit that the categorical distinctions between MSPP, WDM, ROADM, and carrier-class Ethernet are blurring. For sales and marketing purposes, “we have categorized the optical capabilities within WDM, while the other functionalities in support of Ethernet Layer 1 and 2, ATM, and other functionalities we categorize under MSPP-even though it’s the same product that does both functionalities,” explains Ehud Bejerano, product line manager in ECI Telecom’s Optical Networks Division.
Of course, one might argue that the integration of next-generation SONET/SDH, Ethernet, WDM, and ROADM functionality pushes the MSPP into God Box territory. Not so, claims Bejerano, who notes that the MSPP is not trying to be all things to all people. It is merely adapting to service layer changes. “There is a clear migration in the services that the transmission network must provide,” he asserts. “From voice telephony, you go to voice over IP. From plain high-speed Internet access, you go to triple play. These trends are happening not in the transport layer but in the service layer. The transport layer has to adapt. We’re not talking about the God Box,” he says. “We’re talking about a platform that efficiently supports those evolving services in the transport or transmission layer.”
Chris Rivera, director of product marketing in Cisco System’s Aggregation and Access Business Unit (www.cisco.com), agrees that the services have evolved, but he maintains that today’s MSPPs have not strayed far from their original purpose. Rivera arrived at Cisco via its 1999 acquisition of Cerent, the company often credited with creating the first MSPP. Cerent developed the MSPP as “a system architected to be able to exist in today’s SONET and SDH networks,” he recalls. Even though the network is undergoing what he calls “a multiservice transition driven by broadband,” the MSPP has stayed true to its roots.
Griliches also believes today’s MSPPs are exempt from the God Box category. When we deal with legacy traffic, whether it’s ATM, SONET or newer Ethernet traffic, we want to be able to do it all on the best type of [switch] fabric, she says, so in a certain sense, the MSPP has become a bit of a God Box. “But the idea behind the hybrid [switch fabric] architectures, which I think are going to have a more successful play, is that you can put in any interface you want, whether it be ATM, SONET, or Ethernet, and [the MSPP] has a fabric that can handle it. That’s the key.”
The folks at Lucent (www.lucent.com), meanwhile, regard ROADM technology as the proverbial line in the sand between the MSPP and the God Box. “When you start bringing ROADM and high-functionality optical management into those devices, that’s where we believe you cross over the border into that realm of the be-all, end-all God Box,” says Ken Wirth, president and general manager of Lucent’s Multiservice Media Networking Solutions Division. “And you build in complexity and sophistication that 1) customers don’t know how to implement, manage, and operate in their networks, and 2) shifts into a whole different price/cost profile.”
Tellabs (www.tellabs.com) takes an altogether different approach, integrating the MSPP as a blade in its 7100 DWDM system-a system that also includes ROADM functionality. According to Tellabs’ group marketing manager Mike O’Malley, the company has taken a system-level view, integrating MSPP, Ethernet, and ROADM functionality onto a single optical platform. “Is all that functionality an MSPP? I would say the answer is probably no,” he admits. “The basic MSPP would be the SONET functionality plus Ethernet.”
O’Malley reports that the system-level approach resonates with carriers, an opinion corroborated by Verizon’s recent selection of the 7100 for use in its FiOS FTTP network. “We’re seeing carriers looking for a device that can handle the basic SONET traffic, that can handle Ethernet switching. But also, they are interested in ROADM and the ability to optically switch those wavelengths around the network. All of those together really drive that next-generation network build-out,” he says. Because the MSPP is simply one blade in a larger system, it is neither a God Box nor, as O’Malley puts it, “an MSPP on steroids.”
Tellabs’ approach brings up a critical question: Now that the vendors have the ability to construct MSPPs on a blade or chip, what does this mean for the standalone MSPP market? When asked whether the MSPP market was reaching its peak, representatives from Cisco, ECI Telecom, and Lucent answer with a resounding no-for now, anyway. They all acknowledge that growth in the access network will force carriers to evaluate alternative architectures as well.
According to Wirth, Bell Labs just completed a study in which it looked at carriers’ shifting technology and service patterns and attempted to answer the question, “What key trends will occur as operators begin to offer more video services?” Before the Internet boom, carrier networks carried 90% voice traffic and only 10% data traffic (mostly ATM and Frame Relay), recalls Wirth. The Internet boom caused a seismic shift in that ratio, with data accounting for 90% of all network traffic and voice just 10%.
“Now think about the amount of video traffic that will start to come into the network as more customers like Verizon with their FiOS offering and Telefonica with their IPTV [offering] start adding standard definition [TV], which is 2 to 4 Mbits/sec, or high-definition [TV] at 15 Mbits/sec,” says Wirth. “That just throws a tremendous amount of data bandwidth into the network. What I believe is going to happen, and the Bell Labs study supports this hypothesis, is that because everyone is going to video-whether it’s mobile video or video in your home with IPTV-90% of the traffic in 3 to 4 years is going to be video. And data is going to be 10% with voice being a small fraction.”
While this growth will drive additional demand in the MSPP market, it will force carriers to look at other architectures, too. The beauty of the MSPP is that it allows carriers to maximize their return on investment on their existing infrastructure, says Rivera, “but at the same time, broadband services are driving bandwidth demand faster than it’s ever been driven in the past. And the economics begin to prove that you need to take those broadband services that are coming from DSL aggregation and PON systems, put them into a switched packet-based carrier Ethernet infrastructure, and map that over WDM as soon as possible,” he explains. IPTV and triple-play services are driving carriers to build pure packet-over-DWDM or packet-over-optical networks, and this transition could happen sooner rather than later, he says.
ECI Telecom admits that some of its customers may go the dedicated Ethernet route, particularly those in densely populated metros with a greater penetration of broadband services. But in rural networks and medium- and small-sized towns, “the existing SONET/SDH network is perfectly capable of handling those services,” says Bejerano.
“The key thing is not to look at the growth of Carrier Ethernet or a pure Layer 2 implementation coming at the expense of MSPP,” adds Greg Smith, senior manager of product and systems marketing, optical networking, at Cisco. “It’s really going to come down to different options, different flavors. There will be a lot of coexistence for quite some time.”
Griliches points out that there are two prevailing schools of thought among carriers when it comes to metro network design. Those who adhere to the first are the SONET proponents, primarily the incumbent carriers. They understand SONET, they are comfortable with SONET, and they are loathe to walk away from SONET’s reliability. In the other camp are alternative service providers, such as the cable multiple system operators (MSOs), who believe that mapping Ethernet directly onto WDM offers the cheapest and least complex infrastructure.
“It’s funny,” says Griliches, “because it all comes down to the human factor, not necessarily the technology that’s available. It’s just a matter of how comfortable [carriers] are with IP versus the comfort of SONET. They tend to stay in either the SONET world or the IP world, even though those worlds are converging. That’s why there are so many products that allow you to do either/or and both.”
Tellabs’ O’Malley maintains that his company has made the right choice to implement MSPP technology as a blade in a larger DWDM system. While he notes that the international SDH-based MSPP market remains healthy, the U.S. SONET-based market may be reaching its peak. “In the United States, where we’re seeing larger networks and a lot more emphasis on convergence around this idea of the big carriers delivering video, there is now a driver for a high-capacity converged optical platform, which really drives you to DWDM,” he explains. “Based on that, now you’re seeing a lot of additional functionality being combined into a single optical platform. That’s why, from a North American perspective, you’re seeing at least a plateau if not, moving forward, a decline in MSPPs.”
Griliches agrees. MSPP as a TDM-based switch fabric will sell well this year, but “it’s probably at its cap,” she admits. She believes that the market will continue its migration from a TDM-based fabric to some sort of hybrid switch fabric that supports wavelengths, legacy interfaces, and Ethernet. “Call it a ROADM, call it an MSPP, call it metro WDM,” she says. “But it has to support wavelengths, it has to support Ethernet aggregation, and it has to support SONET on a chip. The carriers will figure out what the best architecture is, but it’s not necessarily going to be all TDM based.”
Meghan Fuller is senior editor at Lightwave.