FEBRUARY 13, 2007 By Stephen Hardy -- The Physical Link Layer Working Group of the Optical Internetworking Forum (OIF; search for OIF) will evaluate "alternative" modulation formats to help the ITU Telecommunication Standardization Sector (ITU-T) develop a 10-km interface specification for 40-Gbit/sec transmission. The working group hopes to recommend a single modulation format to the ITU-T by this time next year.
Optical transmission at 10 Gbits/sec and below traditionally has relied on non-return-to-zero (NRZ) and on-off keying (OOK) modulation formats. However, 40-Gbit/sec transmission is significantly more sensitive to such impairments as optical attenuation, polarization mode dispersion, and, particularly, chromatic dispersion. The use of alternatives to NRZ has generally come to be regarded as the best way to support uncompensated 40-Gbit/sec (and higher) transmission rates over currently deployed optical infrastructure.
The ITU-T is working on a follow on to the interfaces contained in Rec. G.693, "Optical Interfaces for Inter-Office Systems," which defines support of 40 Gbits/sec at distances of 2 km or less at 1550 nm using NRZ modulation. The new work would extend reach to at least 10 km by accommodating greater than 200 psec/nm of dispersion. The ITU has asked the OIF to help develop the specification by evaluating alternatives to NRZ, a gauntlet the OIF decided to take up at its most recent meeting in Milpitas, CA, on January 22-25.
According to Karl Gass, Physical Layer User Group chair, and Jim Jones, technical chair, the Physical Link Layer Working Group will examine a variety of potential modulation formats such as but not limited to optical duobinary, differential quadrature phase shift keying (DQPSK), and PAM-4 with an eye toward offering the ITU a single modulation format for its specification.
"We'll come up with a recommendation and we'll provide the data necessary to back up our recommendation," says Gass, who works at Sandia National Laboratories. "Typically we do try to target one answer. That's their preference -- if everything's a standard, then nothing's a standard. But, on the other hand, there are tradeoffs that are made. And we'll try to specify that in 'this situation,' perhaps 'this' is a better answer."
Gass and Jones say that the group's recommendation to the ITU-T will not include a quantitative cost component. However, one can expect that the working group, which comprises component and systems vendors with potentially conflicting interests in the outcome, will have the potential cost of the alternatives in the back of their minds during the evaluation process.
"As a collection of component and system vendors, we're market driven. So we're very sensitive to the cost dimension," explains Jones, who works at Alcatel-Lucent. "As a result, we take that into consideration -- although there's no explicit cost-related data that is quantitatively taken into account in the work. Because of things like our anti-trust policy, we specifically avoid any discussion of price-related items. But, on the other hand, the vendors are sensitive to cost and aren't going to bring in recommendations that aren't going to be market feasible."
Gass points to the fact that it would be difficult for representatives of competing technologies to agree on prices as another factor in staying out of giving the ITU-T specific cost numbers. "Generally, we stick to measurable things," he says.
Availability will have a significant impact on a technology's perceived value, Gass goes on to explain. "The difficulties come in if you a solution for larger amounts of dispersion, is it a proprietary solution?" he says. "Do we have multiple ways to solve that? If we have multiple ways, we have no problem going forward with that. But we don't go forward with a recommendation based on a single-company solution."
While the ITU-T has targeted 10-km reach, Gass expects the group will discuss longer distances. "We will have members push that number because of a business agenda. So we will provide [the ITU-T] with answers past that," he predicts.
Those answers will not include information on pairing the modulation formats in question with either optical or electrical dispersion compensation techniques. Gass says that there is ongoing work within the OIF that covers receiver technology that encompasses electronic dispersion compensation. The new project will limit itself to transmission technology for uncompensated performance, he says.
The group meets quarterly, with the first meeting to tackle the question in depth scheduled for April 24-26. The goal, say Gass and Jones, is to have a draft recommendation by the conclusion of this year's fourth quarter meeting and to have a final recommendation to the ITU-T by the first quarter of next year.
In addition to evaluating the 40-Gbit/sec modulation formats, the OIF also decided in January to start a new project within the Software Working Group, called the "Extension for the Interface Management API" project. The goal is to amend the existing Rev. 3.0 implementation agreement to support features added to Ethernet, SONET/SDH, and PDH by recent standardization work in the IEEE, ITU-T, MEF, and OPTXS.
Visit the OIF