The next revolution in fiber optics for telecommunications networks may be under way in UCLA electrical engineering professor Ming Wu's Integrated Photonics Laboratory.

Wu and electrical engineering graduate students Thomas Jung and Hyuk-kee Sung have fabricated an optical injection system for semiconductor lasers that is housed on a single chip. The system, which can be operated almost like a single laser, offers nearly three times the bandwidth capacity of single semiconductor lasers, enabling an increase in the amount of information that can be transmitted via fiber optic cables.

Supported by the Defense Advanced Research Projects Agency's RFLICS program, the UCLA team has been working with researchers from Multiplex Inc. in New Jersey to develop and fabricate the system. Multiplex produces high-end photonic components for optical telecommunications networks.

An optical injection system boosts the strength of a signal by directing the light from a master laser into the beam of the slave laser. Semiconductor distributed feedback lasers are essential to reliable large-scale optical telecommunications networks.

"This breakthrough extends the application range for semiconductor lasers," notes Wu. "We call it a semiconductor laser on steroids because its performance is so strong."

Although the chip - at 750 micrometers in length - is twice as long as single-laser chips, it can fit in the same packaging that houses current laser chips.

Jung presented the team's findings at the Optical Society of America's Conference on Lasers and Electro Optics earlier this month.

The last major breakthrough in the field came seven years ago, when investigators discovered they could improve the bandwidth and stability of semiconductor distributed feedback lasers by injecting a significant amount of light from a master laser into the modulated laser.

This process, called strong optical injection, can expand the available bandwidth provided by semiconductor lasers by three to four times, enabling greater amounts of complex information to be transmitted more quickly.

The problem, according to Wu, is that current optical injection systems designed to meet the future demands of advanced telecommunications networks require two separate lasers, doubling the cost of the system. The systems also require a physical barrier between the two lasers to ensure one-way injection, which further increases the complexity of the system and limits the impact of the second laser on the first. Strong optical injection systems have thus far been limited to laboratory use, as the operating wavelengths and polarization of the two lasers must be carefully controlled and matched.

What Wu and his team have done is to integrate the two lasers onto a single chip, eliminating the need for an optical isolator. The two lasers are built along a common waveguide and use a grating that is three to four times stronger than normal distributed feedback lasers to confine the light in each laser. The lasers' operating wavelengths are matched by controlling their bias currents. The team's breakthrough has opened the door for more powerful telecommunications systems that will cost significantly less.

"Our next goal is to design and manufacture a chip that will support higher bandwidth systems - 40 gigabits/second systems instead of 10 gigabits/second systems for both analog or microwave fiber optic systems," explains Wu. "Although there are still some unknowns, we are very pleased with our results to date. The chip enables a significant jump in bandwidth capacity and will be much easier to integrate than separate injection systems."

For additional information on research in Wu's Integrated Photonics Laboratory, please visit http://www.photonics.ucla.edu/.

 

-Marlys Amundson 07/09/03