A number of issues arise when one considers implementing a tunable laser in an optical network. First is integrating the tunable laser into the installed equipment base. Today most metro and regional DWDM system providers are migrating to MSA-compliant transponder modules. Numerous vendors are competing in the 10-Gb/second DWDM market for intermediate-reach (IR) through long-reach (LR) transponder modules. Costs have been driven down and performance is continually improving. Adding full C or L band tunability to these MSA-compliant modules is a challenge. Size and power consumption are critical parameters that must be addressed. If the tunable laser cannot fit within the MSA requirements, then a new specification will have to be drafted and existing line cards will have to undergo revision.

The second major issue is the maturity of the technology. Any new technology goes through a classic adoption cycle starting with a few progressive early adopters leading the way. This phase is usually followed by a lull as the market assesses the commercial viability of the technology. If this hurdle is successfully passed then the mainstream customers get onboard — demand goes up and costs come down. Today, full C or L band tunable lasers have progressed past the infancy stage and have reached the point where the early adopters are starting to get serious — qualification tests are being run and product road maps are being outlined.

During this period of technology assessment, an interim path exists for mainstream customers to gain some of the benefits of full C or L band tunability without the inherent risk. This involves tuning DFB lasers thermally. DFB lasers use a thermo-electric cooler (TEC) to maintain wavelength accuracy. Changing the temperature of the DFB laser changes the wavelength. By taking advantage of this feature, an 8- channel tunable laser (50 GHz channel spacing) can be easily derived from existing single channel DFB lasers. While this approach falls far short of full C or L band tunability, it does decrease sparing requirements by a factor of four which is still a major benefit to carriers.

In addition, DFB laser technology is mature and adding the thermal tuning feature will not adversely impact existing MSA transponder modules. An 8- channel tunable MSA transponder module can be a drop-in replacement for today's single-channel design.

As we move toward the long-term goal of full C or L band tunability, optical network equipment designers should look at the benefits of taking an interim step. Thermally tuned DFB lasers offer proven technology, significant network cost savings and drop in replacement with minimal technical risk and little price penalty. Taking this interim step will ultimately reduce the long-term hurdle of integrating full C or L band tunability into optical networks.

John Stewart, is director of product marketing, Transport Products Group, at Vitesse Semiconductor Corp.