NOVEMBER 5, 2008 By Meghan Fuller Hanna -- Ekinops (search for Ekinops) tomorrow will unveil its strategy for 40G transport. Dubbed Wave Bonding, the patent-pending technology enables Ekinops to transport both native 40G and 4 x 10G payloads.
For its part, Ekinops believes there are few applications that require a 40G serial pipe today; instead, the demand will continue to come from the 10G market. And that idea has helped shape the company's 40G offering.
"If you want something that is going to help you with what we think the number one requirement is--which is increasing the number of 10-Gig waves in the C-band spectrum-- you can't go down the path of using something like DQPSK or a technology that, transponder for transponder, is going to be more expensive," explains Rob Adams, vice president of global marketing for Ekinops.
He says there are three key reasons why the industry has not ubiquitously adopted 40G�and they are all tied to cost. First and foremost, a 40-Gbit transponder is not less than four times the cost of a 10-Gbit transponder, says Adams, particularly if you use advanced modulation schemes like differential quadrature phase-shift keying (DQPSK) or differential phase-shift keying (DPSK) as some of Ekinops' competitors have done.
"Even if you take an existing 10-Gig transponder and you compare that to a 40-Gig transponder--just transponder for transponder, forgetting all the other miscellaneous costs that you run into for 40 Gig--because those techs are brand new to the market, [our competitors] require new modulators, new lasers, all sort of new components," Adams notes. "And some of them are introducing new technologies, like Coherent [receivers] in order to push distances a little further. All in all," he says, "it's a bunch of new technologies, and what a bunch of new technologies always means is lots of cost."
Secondly, says Adams, current 40G offerings do not support the same distance as existing 10G systems. It is difficult to overlay 40G on existing 10G routes without disrupting traffic and adding additional regeneration equipment.
Third, polarization mode dispersion (PMD) presents a problem at 40 Gbits/sec. Though advanced modulation schemes like DQPSK improve performance relative to 40G, even these methods require some form of PMD compensation, either as an external compensator or built into the transponder, says Adams, noting that the additional components add both cost and complexity to the network.
Ekinops claims it has solved these issues with its own brand of 40G transport, which will be available on the company's flagship Ekinops 360 platform beginning in early 2009.
"We went down the path of using 10-Gig electronics and optics and essentially very tightly bonding four 10-Gig waves in a form of frequency muxing," Adams explains.
The company says it has developed a patent-pending method for tightly bonding four independent 10-Gbit/sec wavelengths that does not incur any of the non-linear dispersion issues that have caused severe performance degradation in the past. Nor does it suffer from interference between adjacent 10- and 40-G channels, which would require the use of guard bands or empty spaces between the two technologies. As such, Ekinops says its brand of 40G is more spectrally efficient.
"Initially, we believe it will be used as a muxponder to carry more 10-Gig channels in the C-band," Adams confirms. "Other people who are talking about 160, 320 channels--most of them are talking about the L-band, but with this technology, we have no need to go to the L-band. In fact," he adds, "because this technology is really four tightly bonded waves, it's very easy to demux. There's no demodulation. There's no new modulation technology that we have to introduce at the drop sites. Essentially, we just demux it and we don't demodulate it. It just drops off as a native 10 Gig."
Ekinops plans to release two versions of its Wave Bonding technology: In the first quarter of 2009, it will introduce a version that bonds four 10-Gbit/sec wavelengths at 100 GHz, and later in the year, it will introduce a version that bonds four 10-Gbit/sec wavelengths at 50 GHz. (The company is also planning to add a serial interface later in the year, as dictated by demand.)
"At the beginning of the year, even with 100 GHz, you are essentially doubling the number of waves you can carry, taking us to about 160 [wavelengths] or 1.6 terabits on two fibers," Adams says. "And then in the third quarter, we will essentially quadruple 10-Gig capacity to 320 [wavelengths] or 3.2 terabits on a single system with a single pair of fibers."
But perhaps the biggest attribute of its 40G system, in Ekinops' eyes, is its cost effectiveness, thanks to the use of its previously announced T-Chip or Transport on a Chip device. (See "Ekinops touts 'Transport on a Chip' technology"). According to the company, the T-Chip can be programmed for virtually any functionality. The firmware-programmable, field-upgradable chip incorporates all the requisite electronics for transporting 4 x 10G, allowing Ekinops to eliminate three-quarters of the electronics on the board. "When we go to four times the optical capacity, we don't have to go to four times the chips," says Adams. "And it's the same cost as a 10-Gig card in terms of the electronics."
Visit Ekinops