Coherent technology adds spice to 40G alphabet soup

By Stephen Hardy

Overview

A spate of 40-Gbps announcements based on coherent detection at OFC/NFOEC signals yet another technology shift in that space.

If choice is good, life is really good for carriers interested in raising network transmission speeds to 40 Gbps. That’s because a string of announcements from system and module vendors at OFC/NFOEC this past March signals that coherent detection technology for 40-Gbps applications has moved from the exclusive (and popular) domain of Nortel to a more mainstream position. Coherent offerings from multiple systems houses will now battle platforms that rely on differential phase-shift keying (DPSK) and differential quadrature phase-shift keying (DQPSK) for the attention of service providers. And with a three-horse race underway, bets are already being made for win, place, and show market positions.

Coherent detection, as its name implies, isn’t a modulation format. It’s a way of accurately receiving signals—and, in particular, extracting them from large amounts of impairment-induced noise. The secret is in the electronics—analog-to-digital converters (ADCs) that pave the way for algorithms within digital signal processors (DSPs) that aid polarization separation and phase resolution, and clean up of the effects of chromatic and polarization-mode dispersion (PMD). Because of the data rates involved, both the ADCs and DSPs must operate at very high speeds.

The technology works in tandem with modulation formats similar to those that competing systems will employ at 40G. For example, the Optical Internetworking Forum (OIF) has opted to use coherent detection with dual-polarization QPSK in its 100-Gbps framework. So the question at 40G is not whether coherent detection itself has advantages over DSPK or DQPSK, but whether the use of the advanced receiver technology (paired with dual-polarization QPSK) provides enough of a performance improvement over the use of a simpler PSK modulation format alone to be worth the additional cost and receiver complexity.

All aboard the 40G coherent bandwagon

Nortel was the first systems vendor to vote “yes” when it added 40G coherent-enabled capabilities to systems such as the OME 6500 (which was part of the assets Ciena acquired earlier this year). Customers responded well; in the middle of 2009, the Dell’Oro Group pegged Nortel’s share of the 40G market at 41%.

That kind of success naturally caught the attention of competitors. Fujitsu Network Communications (www.fujitsu.com/us/services/telecom/) announced a coherent-enabled 40G option for its FLASHWAVE 7500 ROADM the week before this year’s OFC/NFOEC, and Nokia Siemens Networks (www.nokiasiemensnetworks.com) followed suit during the show with a similar announcement for its hiT 7300 DWDM platform. While no announcement was made, a Cisco ONS 15454 MSTP platform with coherent-based 40G capabilities also appeared on the show floor.

Any platform war needs arms dealers, and it’s no coincidence that these announcements bracketed one from CoreOptics Inc. (www.coreoptics.com). The module vendor unveiled a 40G transponder at the show based on what it called CP-QPSK, which is short for “coherent polarization-multiplexed quadrature phase shift keying.” Both Nokia Siemens Networks and Cisco are announced customers of CoreOptics; while neither company would confirm a connection, it appears likely that CoreOptics has supplied its technology to Fujitsu, as well.

Similar transponders will arrive soon from such companies as JDSU (www.jdsu.com) and Mintera (www.mintera.com), the latter in collaboration with ClariPhy Communications (www.clariphy.com). Sources at Opnext (www.opnext.com) said at OFC/NFOEC that their 40G coherent offering would be available “not too far behind” the 100-Gbps version planned for release by the end of this year.

The fact that the 40G offerings will use the same technology as upcoming 100G transponders enables module vendors to both leverage existing work and, in essence, test it. For example, the same electronics principles apply to coherent detection technology at both speeds—just much faster, at 100 Gbps. Meanwhile, using a 40G offering as an “interim step” toward 100G enables transponder designers to test integration concepts, algorithms, and packaging strategies that might come in handy later. Of course, for those doing 100G first, it seems logical to assume that anything you can do at 100 Gbps you could probably accomplish at 40 Gbps.

Making a choice

But 40G coherent has value beyond a design exercise, as Nortel proved. It offers the most robust performance in the face of PMD among currently available options. However, it carries a comparatively more robust price tag versus DPSK and DQPSK, as well. Service providers will thus face price/performance tradeoffs.

Technology suppliers must make such choices first. For example, Mintera President and CEO Terry Unter said at OFC/NFOEC his company will focus on its version of DPSK (adaptive DPSK) and a coherent option.

“[Adaptive DPSK] allows us to get long reach, very good OSNR performance, very high tolerance to filtering,…50-GHz channel spacing, [and] tolerance to ROADMs,” he said. “The only weak point with DPSK has been its tolerance to PMD in those cases where fiber PMD is high. And that’s led to some proliferation of modulation formats, such as DQPSK, which has better tolerance to PMD. But really not to the level that is fully required. Coherent solves that problem.”

On the other hand—and not surprisingly—Saeid Aramideh, senior vice president of global sales, marketing, and business development at CoreOptics, predicted that coherent would sweep the 40G line-side board. Meanwhile, Finisar (www.finisar.com), Oclaro (www.oclaro.com), Opnext, and Yokogawa (www.yokogawa.com) have confidence in the market pull for their DQPSK offerings.

In the long term, Scott Schube, senior analyst at the market research and analysis firm LightCounting (www.lightcounting.com), likes coherent’s chances. “My take is that everything will eventually go to coherent,” he says. “And when I say, ‘everything,’ I mean everything new, because the old stuff never goes away.” Schube points to the fact that it uses the same technical approach that is the focus of most 100-Gbps efforts, as well as its performance, which he terms “so much superior in so many ways,” particularly in comparison with DPSK.

However, “eventually” isn’t right now. Schube thinks that DQPSK will be the winner over the next two or three years because of the time it will take for more 40G coherent offerings to reach the market and the fact that DQPSK will likely approach cost-parity with DPSK relatively quickly.

So it would appear the alphabet soup of options in the 40G market will continue to boil for several years to come.