Abstract
The pundits of the money world tell us to be a “dotcom” or enable them for excitement and rewards. Traffic on the Internet Highway is certainly stepping up the pace as “slow” electrons make way for ultimate-speed photons creating major hardware opportunities. The “Copper Road” has become the “Glass Super Highway” as long-haul terrestrial and underwater communications links move up to Advanced Photonics.
Nothing can be faster than light, but more important, no other medium can offer wider bandwidth when wave-multiplexing strategies are used. Photonics, employing dense wave division multiplexing (DWDM) can carry the equivalent of 12,000 encyclopedias or 5-million phone calls on a single fiber. Recent advances in photonics hardware, including higher-powered lasers, more efficient amplifiers and cleaner optical fiber are enabling incredible bandwidth for the Internet and general communications services. But how do we route a light beam?
The long-haul segments of the Internet, now mostly fiberoptics, have been converting modulated light to electronic signals, routing with conventional electronic hardware and then re-converting back to light. Yes, O-E-O (Opto-electro-opto) works, but with cost and time-delay penalties. The communications industry has decreed that the double conversion process must go, but what technology will be the replacement?
Enter optical MEMS, or MOEMS (micro-opto-electro-mechanical systems). The MOEMS switch/router approach was endorsed by the Internet carrier and hardware industry that paid billions of dollars in 2000 to acquire MEMS companies, some that had not even shipped a product. But what are the issues and are there competing technologies that could win?
Micro-mirror technology is at the top of the popularity chart right now. Can MOEMS mirror routers solve cost problems and can they even switch at the rates demanded. What is the ideal mirror switch strategy: binary “off/on” or point-to-light pipe arrays? What about other MEMS approaches such as micro-bubble fluid beam refraction that appears to offer a much simpler construction? Maybe the mechanical devices are only an interim destined to obsolescence by a future solid state optics switch. The optical switch, powered totally by photons, is already in the lab and could be the final answer.
This paper will survey MOEMS inside the Internet to seek answers to the billion dollar questions. The focus will be on micro-mirrors and their packaging issues both inside and out. We will deal with selecting the ideal optical MEMS package and choosing the right atmosphere control. Certain in-package contaminants are death to mirrors, but they can be controlled even if generated after the package is sealed. So tune in to find out if MEMS can catch the WAVE!
