Optical Fiber

Annie Lindstrom

The recent market downturn doesn’t mean the major fiber producers have put research and development on hold.

At the beginning of 2001, manufacturers of optical fiber couldn’t make the stuff fast enough. By midyear, the economic boom’s bubble had burst, and the world’s leading suppliers began applying the brakes to their global fiber manufacturing operations. Market demand seemed to diminish overnight after nearly two years of unprecedented growth.

Although fiber makers may have slowed their production temporarily, they haven’t stopped concentrating on improving their product, according to Janice Haber, vice president for systems engineering and market development for the OFS Div. of Furukawa in Atlanta. (Furukawa acquired Lucent Optical Fiber Solution (OFS) last year.) “Work … in research, materials and measurements is still going on,” she said. “We have to stay as far out in front [in research and development] as we can. If we don’t, someone else will do the innovation.”

From the deep blue

Fiber’s evolution in recent years somewhat resembles that of the humans deploying it, albeit within a substantially compressed time frame. Many fiber innovations make their way to terrestrial networks after first being tested and tried in undersea networks. One example is a set of dispersion-managed fibers for undersea applications that some are claiming can more than double the capacity of transoceanic communications systems.

The fiber accomplishes this by combating the effects of chromatic dispersion by using dispersion itself. Previously, undersea systems used alternating lengths of same-slope positive- and negative-dispersion fiber, with the cleanup of any residual chromatic dispersion being the primary responsibility of dispersion-compensation modules in the system amplifiers. Modules are no longer necessary because fiber manufacturers splice two fibers with inversely identical dispersion characteristics, creating an optical fiber that flattens the dispersion slope for hundreds of channels.

The key to making dispersion-managed fiber is to match the negative and positive dispersion slopes “perfectly,” Haber said. This is achieved by using proprietary “profile” controls that are employed to make preforms from which the fiber is drawn and to draw the fiber itself. As a result, the firm’s UltraWave dispersion-managed fiber (Figure 1) can carry up to 64 channels at 10 Gb/s. Such capability is important to submarine system operators because it allows adding capacity to an undersea link without laying more fiber.

Figure 1. The inverse dispersion slope of dispersion-managed fiber cancels the detrimental effect of dispersion slope across a wide spectrum of wavelengths, enabling a dramatic increase in the number of DWDM channels used in ultralong-haul transmission. Courtesy of the OFS Div. of Furukawa.

An elegant solution

Dispersion-managed fiber is, in essence, an elegant solution to the problem of chromatic-dispersion compensation in submarine systems because it “takes care of itself,” according to E. Alan Dowdell, new-products manager for Corning Inc.’s Optical Fiber Div. in Corning, N.Y., which calls its brand Vascade R1000. Another feature is inherent temperature stability; i.e., temperature change has no impact on dispersion because both the negative- and positive-dispersion fibers are subject to the same temperature variations. “It’s a clamped system,” Dowdell said, “so if one effect happens to one part of the system, the other compensates for it automatically.”

Because a dispersion-compensation module is not needed, system design is simplified. One benefit, said Dowdell, is a lower noise figure for the amplifier.

Furukawa is applying some of the ideas used to make the matched-dispersion fiber for undersea systems to its next generation of TrueWave fiber, which was scheduled for unveiling at last month’s Optical Fiber Conference (OFC). The goal with this fiber type is to open the S-band for Raman amplification. “You can put your Raman pump lasers into wavelengths that do not interfere with your transmission wavelengths,” Haber noted. “That opens up more capacity for transmission channels and allows service providers to transmit data 200 to 600 km farther than possible with existing fiber.”

The company is also promoting dispersion-managed fiber for terrestrial applications. Before the market softened last year, Haber said, there was a “strong notion” that the fiber type might be of use in ultralong-haul terrestrial networks. One potential stumbling block, though, involves perfect matching. Although their negative and positive dispersion slopes are matched perfectly, the fibers’ effective areas are not. Therefore, the fibers for submarine systems are spliced at the manufacturing facility before delivery to the cabler. The firm has spent about a year developing field-deployable splicing capability before introducing dispersion-managed fiber as a terrestrial solution.

The next generation of terrestrial systems could be some form of a dispersion-managed system in long-haul networks, Dowdell said. “I think the technology will become viable in terrestrial networks when higher bit rates become more prevalent. People aren’t deploying 40-Gb/s systems today, but it’s just a matter of time. As soon as that happens, you are going to see people take a whole different look at their outside plant.”

High and dry

Another fiber type is just now coming into its own. Lucent Technologies introduced its low-water-peak fiber under the name AllWave in 1998. Corning and Alcatel unveiled their versions of the enhanced single-mode fiber — SMF-28e and Enhanced SMF (E-SMF), respectively — last year.

Standard single-mode fiber absorbs hydrogen in the 1383- to 1480-nm band. The resulting water peak degrades fiber attenuation and blocks the use of about 30 percent of the fiber. The low-water-peak fiber is identical to the standard type in every way, except that hydroxyl ions are removed during the manufacturing process (Figure 2). The absence of that hydrogen opens up 100 nm of previously unavailable spectrum to service providers.

Figure 2. Low-water-peak fiber removes attenuation contribution by OH^— and opens up the E-band for coarse WDM transmission. Courtesy of Corning Inc.

“With enhanced single-mode fiber, you can operate across the whole region from 1260 to 1625 nm,” said Jim Ryan, global product manager for Alcatel’s Fiber Div. in Claremont, N.C. Until recently, however, that capability has been of little use to service providers. The reason is simple. Because the standard fiber did not make use of the previously blocked spectrum, equipment was not designed to do so.

Now that the situation has changed, equipment suppliers are introducing transmission gear such as coarse wavelength division multiplexing (WDM) equipment, which uses wider channel spacings than dense WDM equipment and exploits the newly opened window. Furukawa and LuxN Inc. promised to demonstrate the coarse transmission gear at OFC. LuxN is among “at least half a dozen equipment suppliers” sending out teasers that they may roll out such equipment this year.

In addition to the low water peak and low attenuation (0.22 dB/km) at the 1550-nm wavelength, Alcatel’s enhanced fiber specifies a polarization mode dispersion value of 0.08, which is less than the 0.1 specified for the standard type. Although only 20 percent lower, this spec extends the reach of the fiber by 50 percent, Ryan said. “That small change will significantly impact the efficiency of 40-Gb/s transmission systems.”

Metropolitan dark-fiber provider American Fiber Systems and other service providers are beginning to deploy the enhanced single-mode fiber in their networks. “We use low-water-peak fiber because it is one step better than the standard type,” said Kevin Mullaney, the company’s chief technology officer. “If we are going to put in special fiber, we want it to address both today’s needs and tomorrow’s.”

Although the Rochester, N.Y.-based company has paid a premium for the low-water-peak fiber in the past — the firm began deploying it in 2001 — the increase in suppliers means more competitive pricing compared with standard single-mode fiber. So far, the firm has deployed the enhanced fiber in four of the five markets in which it operates.

American Fiber does not light the fiber it deploys. Instead, it sells dark fiber to a variety of customers with diverse requirements. Versatility is the reason the company prefers the enhanced single-mode fiber to the metro non-zero dispersion-shifted option, which is optimized for DWDM in the 1550-nm band. Although Mullaney doesn’t know if any of his customers have lit the 1400-nm window yet, “they can if they want to,” he said. He believes that eventually most service providers will choose low-water-peak fiber over the standard type.

Big ring

On the other hand, Alcatel, which launched its Teralight family of non-zero dispersion-shifted fiber in 1999, recommends that service providers take into account the total length of the rings they are building before dismissing the need for the fiber type in metro and regional fiber rings.

“In the metro environment, service providers will be going to wavelength routing soon, so the systems will not be regenerating signals at each node,” Ryan said. “Wavelengths will be traversing the fiber all the way around the ring, so the total distance of the ring will become the limiting factor.” He believes that, when total ring distance is longer than 80 km (Figure 3), the non-zero dispersion-shifted option will offer an advantage because of the dispersion limits of standard and enhanced single-mode fiber when it is used with 10- and 40-Gb/s transmission systems.

Figure 3. From 80 km to the point where you reach the polarization-mode-dispersion limit of 40-Gb/s transmission systems, non-zero dispersion-shifted fiber (NZ-DSF) effectively lengthens the reach of such systems by almost five times that of standard single-mode fiber (SSMF). Courtesy of Alcatel.

Corning management appears to agree. Dowdell reported that customers have been installing two of its non-zero dispersion-shifted fiber brands in applications where large metro and long-haul networks traverse core metropolitan networks. The company lowered the polarization mode dispersion of its fiber by closely controlling the circularity of the core of the fiber using a technique called outside vapor deposition. During the process, a silica-germania compound is deposited onto a spinning preform. “Since it’s spinning, it inherently generates circular glass,” he said.

Besides delivering improved polarization mode dispersion, the deposition process is very scalable. “Even lower dispersion is better,” Dowdell said. “I don’t want to mislead you and say we have solved the problem.” Although this parameter is controllable to a nice level for communications fiber, the industry is focusing on lowering such dispersion in components. Fiber in amplifiers and dispersion-compensation modules actually contributes a significant amount of polarization mode dispersion to the network.

The future

Overall, the big-three fiber suppliers are focusing on the future by working to ensure that the fiber they develop is compatible with the 40-Gb/s systems. They don’t seem to feel comfortable, however, with predicting just when those systems will become practical reality.

“Our focus is on where transmission systems’ capabilities are heading,” Ryan said. In the metro market, the goal is to provide the capability to deal with wavelength routing and transparent rings. In the medium to long haul, the emphasis is on longer distances with higher bit rates and the ability to properly manage the chromatic and polarization mode dispersion of all the wavelengths, especially at 40 Gb/s.

Meet the author

Annie Lindstrom is a freelance telecommunications writer based in Cape Coral, Fla.