Standards committees examine 1600-nm window

By William B. Gardner

The possibility of a fourth fiber-transmission window around 1600 nm was discussed in this column three years ago (see Lightwave, Aug. 1995, p. 24). Although no standards body has yet approved the use of this window, it is already used for monitoring in a number of commercial systems. Optical Supervisory Channels (oscs) have been proposed for 1625 nm. Recent laboratory improvements in optical amplifiers have extended the useful range of the amplifiers beyond 1600 nm. These developments, plus the continuing demand for more optical bandwidth, have stimulated new studies of optical cable attenuation in the fourth window.

The concerns of fiber and cable suppliers are chiefly the low temperature (-40°C) increase in cable attenuation and the macrobending loss in fiber loops in splice cases. Both of these attenuation increases may become larger as the wavelength increases. Since oscs operate at only a few megabits per second, they can tolerate more loss than can the gigabit-per-second signal channels. Recent studies by several vendors indicate that most currently produced fibers and cables can meet the attenuation demands of oscs. The same holds true for non-zero-dispersion shifted fibers (nzdfs--for a discussion of these fibers, see Lightwave, Dec. 1996, p. 62). Although many conventional dispersion-unshifted fibers can also meet the attenuation needs of oscs, these fibers, with their dispersion zero near 1310 nm, have several times more chromatic dispersion in the fourth window than nzdfs. Compensating out this large dispersion could be a problem for conventional singlemode fibers in the fourth window.

At its June meeting in Vancouver, BC, Canada, the Telecommunications Industry Association (tia) Subcommittee FO-6.7 on Fiber Optic Cable, chaired by Eric Loytty of Siecor Corp. and co-chaired by Mike Kinard of Lucent Technologies, examined these issues. They concluded that osc and monitoring applications should be feasible in the fourth window of currently produced fibers and cables. But this conclusion does not necessarily extend to the embedded plant, since much of it was installed at a time when 1600-nm usage was not contemplated.

To open the fourth window for use, nearly all lightwave standards documents will have to be modified. Presently, the International Telecommunication Union (itu) fiber Recommendations (such as G.652 and G.655), system Recommendations (such as G.957), and component Recommendations (such as G.671) give the "maximum wavelength of anticipated use" as 1580 nm, or in some cases, even less. Nippon Telegraph and Telephone of Japan has proposed raising these limits already. tia`s FO-6.7 has agreed to work toward increasing the limit to 1625 nm. Up to this wavelength, optical time-domain reflectometers are readily available, and they will be a crucial tool for characterizing fourth-window attenuation in both the factory and field. This U.S. position will be carried to October 1998 meetings of both the itu and International Electrotechnical Commission.

Other changes must follow. The current 1560.6-nm upper limit on the wavelength grid in itu`s Recommendation G.692 would have to be increased. Maximum limits on fiber chromatic dispersion will have to be re-examined. tia FO-6.7 agreed that more data is needed before setting specifications on fibers and cables in the 1600-nm region. q