Optical network developers strategize for a return-to-profitability

Optical network developers strategize for a return-to-profitability
In conjunction with the Optical Fiber Communications (OFC) conference taking place in Atlanta, contributors to this week's In Focus take a look at the latest optical components for implementing photonic transport services. With the over-capacity of bandwidth in the optical backbone that resulted from the technology surge of the 90's, this kind of work optimizes that capacity, but also makes economic sense in the current severe downturn in the communications industry.

The basic strategy: to leverage the existing infrastructure with components that create more services over existing fiber. New approaches to wavelength switching and routing offer more flexibility for service providers to stimulate demand with attractive features. Contributors offer some innovative solutions that may push optical services forward, including excerpts from several research papers at OFC.

Economics is playing a major role in determining the direction of optical research. Across the board, optical network companies are searching for components and technology that will add value to existing networks. "The biggest change that we see taking place is high interest in the cost- per- port solutions," said Samir Samhouri, director of marketing for Agere's access and transport division (Allentown, Penn.), which drives the company's optical communications strategy. "When we say "cost", that doesn't necessarily mean that they are looking only for a cheap kind of device, although that does come into play," he said, "but they are looking for a single solution to support multi-protocol, multi-reach, multi-channel applications." A basic cost sensitive strategy is to incrementally add optical capability as a system is deployed to reduce upfront costs as much as possible.

More emphasis is being placed on lower rate solutions this year. "Pluggable 10G solutions will be a big deal from a technology standpoint," noted Samhouri, "but from a real business deployment standpoint, the question is going to be how can you get the most number of ports from a linecard at the lowest cost per port."

Agere has built a Sonet SDH framer that can be used on all optical line cards from OC-3 up to OC-192. The advantage: one platform can be used everywhere and companies do not have to use a different device for each of the linecards. "This impacts their cost and design expenses because they don't have to replicate their design effort — which is significant in the current market conditions," said Samhouri. He said that Agere's overall strategy will be "any service, any port at the lowest cost and power — that applies from a copper access standpoint as well as an optical standpoint."

Chris Hamilton, segment marketing director in systems architecture at Agere, co-authors a contribution to this week's In Focus section on framer ICs for the Sonet/SDH using virtual concatenation and generic framing protocol (GFP). Hamilton along with colleague, Nathaniel Grier, discuss the realities of designing in an environment where "the days of "build it and they will come" has abruptly ceased in the communications world." Their team has integrated the capabilities of traditional framers with companion devices to enable single ICs and single OEM designs to support multi-rate, multi-protocol, multi-channel solutions over Sonet/SDH.

Hamilton notes that there are two key issues that suppliers need to address to succeed in the current market: how to make traditional and current networks more profitable and how to return service providers to profitability through their current installed base. The answer, he believes, lies in "being able to have the technology in standard products that allows you to use a single product across multiple platforms and across multiple protocols." He expects that the direction for next generation optical networks will be the ability to develop not only the Sonet traditional network, but to also integrate T-1 and T-3 with Sonet functionality.

Painless integration

Integrating wavelength switching within existing networks is the top priority for Nortel Network's Brian Lavalle, a senior manager in the photonic systems engineering department (St-Laurent, Qubec). "As capital influxes into the telecom industry continued unabated for years, numerous innovative technologies were introduced, several of which were targeted towards wavelength switching applications," he notes in his article. "Rather than switching traffic in the electrical domain, traffic would instead be switched purely in the optical domain. However, most of these innovative technologies were never actually commercialized for technical as well as economic reasons." He argues that cost-effective and relatively painless integration into existing network structures "is absolutely mandatory and expected by carriers."

Aiming at DWDM applications, managers Mike Blake and Linda West of UK-based NP Photonics describe a new MEMs Fabry - Perot filter that uses compliant elastomeric materials to support movable mirrors. The materials are said to be as much as six orders of magnitude less stiff than silicon, and can be deposited in a broader range of layer thicknesses. Unlike carbon-based elastomers, the materials have a Si-O-Si backbone giving them excellent mechanical, chemical and thermal stability. This type of design is ideal for implementing optical crossbar switches for flexible wavelength routing.

And, engineers Helen Stapleton and Brian O'Mara of Analog Devices, Inc. (Limerick, Ireland), explain how the evolution of optical multi-source agreements (MSA) specifications such as XFP, XENPAK, XPAK and SFF-8472 have become a challenge to optical and electronic network designers. Their solution is a single chip reprogrammable data acquisition system that offers control and monitoring services to the designer.

From a research standpoint, several papers from OFC reflect ongoing work that is creating simpler and more effective optical components that might offer cost savings in the next generation optical network. For example, an approach to optical fiber fabrication called microstructure-fiber has sparked an active area of optical research. Rather than attempting to get the most uniform glass structure in an optical fiber, the microstructure approach introduces carefully engineered non-uniformity to create a more optically active medium.

The technique has been successfully applied to Raman amplification, signal regeneration, wavelength conversion and optical switch designs. In their contribution, which will be presented in full at OFC, a group of optical researchers at Northwestern University (Evanston, IL) describe a prototype microstructure fiber optical parametric amplifier (MFOPA), which they say is the first such amplifier to operate in the 1550 nanometer range used by optical networks. In experiments, the amplifier showed a 15dB gain over a 20 nm range with a peak net gain of 22.5dB.

Meanwhile, supercontinuum generation represents an important capability for wavelength- division multiplexing components that has also been tackled both by microstructured fiber and highly nonlinear fiber systems. Usually, a very fast pulsed laser is required to generate supercontinuum output, but research groups at OFS Labs, (Somerset, NJ) and Lucent Technologies Bell Labs (Holmdel, NJ) have discovered a way to produce supercontinuum output over a bandwidth of 247 nm centered at 1483.4 nm, with good long-term power stability. The researchers achieved their results with a continuous wave Raman fiber laser pump operating at 1596 nm. The full potential of the source is then shown by generating a supercontinuum with a bandwidth greater than 247 nm in 4.5 km of highly nonlinear fibers (HNLFs).

Materials, carbon nanotubes

Optical engineers have long sought a materials system that would allow full integration of optical components in the way that silicon has become the medium of electronic integrated circuits. Excerpts from an OFC paper written by engineers at Little Optics, Inc. (Annapolis Junction, Maryland) describes a high refractive index material called Hydex that the company believes will open the door to large-scale integrated optics. Using conventional CVD processes familiar to VLSI fabs, the materials refractive index contrast is adjustable from zero to over 20 percent. Requiring no anneal steps, the material is said to be easily processed to create a variety of optical components and interconnect on a single substrate.

Applying carbon nanotube technology to optical system design is another area of ongoing work. Researchers at Japan's National Institute of Advanced Instruments, Science and Technology, (Tsakuha, Japan) explain how they have built a simple, robust saturable absorber by depositing nanotubes on a substrate. Saturable absorbers are very effect noise suppressors in optical fiber since they preferentially attenuate amplified spontaneous emission noise and pass the higher-power signal component. Normally, complex compound semiconductor devices are needed to achieve the benefits of saturable absorption.

And, a research team at Sumitomo Electric, Industries, Ltd, (Yokohama, Japan) describe a micro-mirror array technique in their OFC presentation for managing chromatic dispersion in DWDM systems. The researchers bounced multi-wavelength signals from an optical fiber off of a bulk diffraction converting it to a parallel stream of wavelength channels that were focused on a micro-mirror array. The moving mirrors can individually adjust each channel's phase, conditioning the signal so that chromatic dispersion is eliminated.