Bend-insensitive optical fibers simplify fiber-to-the-home installations
The dream of installing optical fiber all the way to the home began in the early 1990s after extensive deployment of optical fibers in long-distance and metropolitan area networks.1 However, significant commercial deployment of fiber-to-the-home (FTTH) in North America started only a couple of years ago, led by Verizon. One reason for the slow growth of FTTH is higher installation costs compared to copper networks. Although optical fiber has superior signal capacity and immunity to electromagnetic interference, there is one basic aspect where fiber lags behind copper cable: the so-called ‘bending loss.’ When standard single-mode fiber (i.e., compliant to ITU-T Recommendation G.652, which defines the geometrical, mechanical, and transmission attributes of single-mode optical fibers) is used, the fiber cable must be installed very carefully to avoid small bends along the fiber path, which can cause signal loss. This is mainly a problem with inside plant cabling deployments such as those within multi-dwelling unit (MDU) buildings during which installers encounter numerous right-angle turns. In addition, the bending requirement for deployment dictates minimum dimensions of hardware and imposes significant costs for FTTH installations.
Several approaches have been proposed to reduce the bending loss of single-mode fibers. Many of these advancements have focused on changes to the cladding, a mirror-like sheath that helps contain the light waves within the core when the fiber is bent. Recent approaches include reducing the mode field diameter (MFD),2 depressing the cladding,3 adding a low index trench, 4,5,6 and adding a ring of symmetric holes within the cladding.7,8 Figure 1(a–d) shows schematics of these fiber designs. Among them, hole-assisted fibers, Figure 1(d), offer superior bending performance, but they are not compliant with ITU-T Recommendation G.652. In addition, the process for producing hole-assisted fibers is much more complicated than conventional fiber-making processes, making the fibers less attractive for large-scale and cost-sensitive FTTH implementation. Although conventional fiber design approaches deliver fibers that meet the standard requirements, their bending performance needs further improvement to meet the demanding requirements of low-cost FTTH installations.
Recently, we have developed a new fiber design technology, called nanoStructures™, for making optical fibers. This is a breakthrough technology that adds new dimensions to the conventional fiber design space. This technology enables new fiber designs with superior bend performance that meet the FTTH requirements and, at the same time, are compatible with large-scale manufacturing and field installation procedures. Figure 1(e) shows a schematic of our fiber design, which consists of a germania-doped core and a nanoStructures ring within the cladding. This fiber design consists of engineered features in the range of a few nanometers to several hundred nanometers.
This design offers several advantages compared with other technologies. First, the refractive index dependence of nanoStructures glass is very different from that of glass with conventional dopants used in fiber manufacturing. This refractive index has much stronger wavelength dependence than that of fluorine-doped glass. This dependence is explained schematically in Figure 2. Light at a longer wavelength sees more of the nanoStructures features than at a shorter wavelength. As a result, the average index change for these features increases with wavelength. This feature maximizes bend performance in the 1550nm window while maintaining a cable cutoff wavelength below 1260nm. Second, large, negative index changes can be made with our design. A relative index change as high as several percent can be achieved by using this new design. Such a high index change is very difficult to realize using the conventional fluorine-doping technology. Third, the scattering property of nanoStructures glass also has strong wavelength dependence. Light at shorter wavelengths has higher scattering losses than at longer wavelengths, which facilitates the tunneling of higher order modes through the ring. These new features can be used to design fibers with much better bending performance while other optical parameters stay compliant with the standards.
We made nanoStructures fibers using conventional outside vapor deposition (OVD) equipment. Fiber results demonstrate that this technology is compatible with the OVD process and suitable for large-scale production. Figure 3 compares typical bend losses as a function of bend radius for the nanoStructures fiber, standard single-mode fiber, and other bend-tolerant fibers at 1550nm. This figure shows that the nanoStructures fiber has about 500 times lower bending loss than the standard single-mode fiber, 100 times lower bending loss than the depressed-cladding fiber, and approximately 10 times lower bending loss than the trench fiber. The typical bending loss at a 5mm radius is 0.03dB/turn.
Other typically-measured optical properties of the fiber samples are summarized in Table 1. The optical parameters shown in Table 1 are fully compliant with the G.652 standards. We also evaluated fusion-splicing performance for these fibers. The fusion splicer was a Fujikura FSM 40S, and a splicing-factory-set splicing program was used. For splices between nanoStructures fibers, the average splice loss at 1310nm was 0.029dB. For the splices between nanoStructures and standard single-mode fibers, the average splice loss at 1310nm was 0.033dB. These results are similar to standard single-mode fibers.
In summary, our nanoStructures technology enables new fiber designs with ultra-low bending loss. The excellent bending performance of these fibers lends itself to the demanding installation requirements of FTTH networks in MDUs and makes them the best choice for such applications.
Ming-Jun Li is a research fellow with Corning Inc. His research work is related to new fibers for different applications. He holds 33 US patents and has published one book chapter and authored and coauthored over 120 technical papers in journals and conferences.
