The wave of the future

At least two of the components that will enable the optical networks of the future have found inspiration in technologies not originally developed for communications.

Just more than a year ago, Agilent caused a stir when it unveiled its photonic switching platform, which is based on technology developed for inkjet printers. When an individual wave needs to be moved from one fiber to another, the platform heats a liquid that then forms a bubble. When the wave hits the bubble, its direction is changed.

Similarly, microelectro mechanical systems, or MEMS, were used for some time in fields such as healthcare and electronics before moving into communications. MEMS-based products use microscopic mirrors to direct beams of light from one fiber to another.

Together, the two technologies will account for about 85% of the global optical switching fabric market by 2005, according to The Yankee Group (see figure).

As with any two technologies designed for the same purpose, there are tradeoffs when one is chosen over the other.

For Agilent, one of the biggest selling points of bubble technology is that it has no moving parts.

The lack of long-term data, however, makes this a benefit only in theory, said Karen Liu, co-director of optical components for RHK. “There's not enough experience in the field for people to know what the failure rate of the [bubble-based] switches are. I think there's a general principle that no moving parts is a good thing. There just haven't been enough devices out there.”

The biggest differences between MEMS and bubble technology, said Jay Patel, senior analyst at The Yankee Group, is the ability to scale. The largest port count for Agilent's bubble technology is 32 × 32 but can reach 512 × 512 ports.

MEMS, on the other hand, can reach a much higher port count, with companies currently working on products that support 1024 × 1024 port and talking of going even higher, Patel said.

That scalability probably is going to be an advantage for MEMS in the production of large switches, optical cross connects and optical add/drop multiplexers, Patel said.

Though MEMS has an advantage in scale, bubble technology has a potential leg-up in manufacturing costs because of its basis in inkjet printing, Patel said.

“For bubble, they are able to use the same production line that they have for the printers. They do have the economy of scale, which you cannot have for any other technology,” Patel said.

There is, however, power in numbers, said Conrad Burke, senior vice president of marketing for MEMS component maker OMM. “If you look in terms of MEMS and the number of companies embracing MEMS, that far outweighs the one company that's supporting bubble technology,” he said.

Liu agreed that the number of companies investigating MEMS technologies could be an advantage.

However, because both technologies are fairly new, there is uncertainty about what each is best suited for in the long-term, said Marlene Bourne, senior analyst with Cahners In-Stat Group.

“There's no one technology that's going to take the entire [optical switching fabric] space. Each is going to find a space where they work perfectly well. It's a question of where and when,” she said.