The Elegant Optical Network

Alan Dowdell

Consumer technologies with which we are accustomed, such as the personal computer, have become more complex in their internal workings, while the interfaces have become much more elegant in their simplicity. This dual trend is reflected in optical networks as well.

Highly dense optical circuit packs, multiple network layers, optical switches with thousands of micromachines and more intense network monitoring all will make the networks more complex. On the other hand, weblike interfaces, common control electronics and outside-plant technologies will make networks seem simpler to those who install the fiber and equipment and to those who operate it.

In the past, networks were optimized for specific applications because access, metropolitan, regional and long-haul systems tended to have their own array of technologies and architectures. Today, operators can implement functions more inexpensively across all networks, thanks to several disruptive technologies ranging from novel modulation formats to new methods of amplifying signals employing stimulated Raman scattering.

Another technology, forward error correction, was developed for radio communications from deep-space probes, but in fiber networks it can add several decibels of gain at little cost. In access and metro networks, new electro-absorption and directly modulated laser technologies continue to raise performance and lower costs and, combined with high-speed optics, will help drive the emergence of 40-Gb/s Internet protocol ports.

With these tools at hand, network designers will push for more integrated architectures that will allow simpler operations, fewer spare parts, less training and reduced costs overall. Their networks will be able to move massive amounts of data quickly and easily — necessary because the margins for transporting data are so much less than those for moving switched voice packets.

Fiber selection

The past three years have been a wild ride in the optical fiber industry. Most of the excitement has been in the long-haul market as a “land grab” mentality of fiber build-outs that began in North America in 1998 was followed by the pan-European builds in 2000. These built-out regions will experience a continued meshing of optical path topologies as well as installation of new technologies that will help stabilize long-haul fiber deployment growth.

A number of other projects were sprinkled throughout Latin America, the Asia/Pacific region and Eastern Europe, but the focus of recent months has been on China as new entrants and recently divided incumbents have begun the daunting task of installing high-fiber-count cables along the eastern seaboard and into the central provinces.

In the near term, operators deciding which optical fiber to deploy will turn to diverse sets of solutions.

In one, optimized performance required by specific sections of the network will determine the fiber type. Enhanced multimode fibers will continue to be deployed at the premises level, for example, where low-cost vertical-cavity surface-emitting lasers are the primary driver. In the ultralong-haul area, where data traffic is aggregated into cost-efficient bit rates and channel counts, fiber selection will be based on achieving the highest-performance, most cost-effective and upgradable optical solutions.

If performance, cost-efficiency and upgradability drive fiber selection in ultralong-haul applications, the criteria for fiber in access, metropolitan and regional networks will be low attenuation, good splicing capability, excellent handling and a dispersion profile suitable for multiple applications.

Meanwhile, submarine networks have moved dispersion management — that is, managing the slope of dispersion — into the fiber plant itself. Elegant dispersion-managed fiber technology is proving itself there, and it is only a matter of time before it is deployed in the high-data-rate segments of the long-haul network.

A separate set of fiber solutions will find application in access, metro and regional networks — the largest growth segments. To maintain a simple, single network across all of these segments, outside-plant personnel will require equipment that is easy to install, maintain and operate. The technologists will demand high-performance fiber and components that will deliver greater capacity and enable simple upgrades and low transport costs.

The major criteria for fiber in access, metro and regional networks will be low attenuation, good splicing capability, excellent handling and a dispersion profile (with dispersion compensation) designed for multiple application spaces.

It is difficult to predict trends beyond these imminent fiber network advances, but deployment of some technologies clearly will not happen in the near future. For example, although much research has been focused on photonic crystal, or “holey” fibers, they will not be transmitting signals in commercial networks for at least five years. Nor will newer materials begin to replace silica-based optical fibers. Lastly, the debates on transport rates and protocols will probably continue beyond 2006.

Although we have completed an exciting time in optical network build-outs, there is good evidence that much more excitement is ahead. Refinements in the technology thus far have only scratched the surface. Clearly, designers and suppliers alike must be prepared to spend the time and energy necessary to deliver the right selection of components in the right combination and packaging to enable the next generation of elegant optical networks.

Meet the author

Alan Dowdell is new products manager for Corning Inc. in Corning, N.Y.