Optical Injection Locking of VCSELs
Optical injection locking uses a second (master) laser to inject photons at a similar wavelength into the transmitter laser (termed follower or slave). Under certain conditions, the follower laser is locked in phase to the master, and the laser dynamics fundamentally change and can result in far better device performance.
With this technique, a directly modulated laser can be upgraded by using a second low-cost CW laser as a master. Experiments to date have implemented this by using discrete components, and either using a fiber optic circulator to couple the light, or by accessing the back facet of an edge-emitting laser. The enhancements that we have demonstrated in VCSELs include: a reduction of frequency chirp (by 3X); a resonance frequency increase from 4 GHz to nearly 40 GHz, yielding a record intrinsic VCSEL bandwidth; a huge modulation efficiency enhancement by > 10 dB (i.e. RF gain); a laser noise reduction; and a reduced laser distortion (by up to 35 dB).
The terrific performance demonstrated by these lasers needs to be studied further, and paves the way to further research such as the integration of the two lasers and making arrays of locked transmitters. Ultimately, directly modulated diode lasers, which are compact, low cost, and low power consuming, have the potential to be applied for 10-40 Gb/s systems, as well as analog RF links.
Research on injection locking will include efforts in modeling, device design and fabrication, experiments, and network integration. Many issues are yet to be resolved before high performance transmitters using injection-locked VCSELs can be accomplished and this technology is used in communication systems.
First, experimental work will continue to seek even higher performance for single devices, with VCSELs designed to provide improved intrinsic modulation characteristics. For these devices, high speed packaging for 50+ GHz will be developed, and we will investigate the use of new nano-structured materials, such as quantum dots and photonic crystals.
For this technology to make an impact, monolithic or heterogeneous integration of the master laser with the VCSEL needs to be accomplished. Integration will be pursued by several approaches including vertically integrated VCSELs, horizontally integrated VCSELs, and the integration of an edge emitter with a VCSEL. The development of thin-film optical isolators for the structures will be considered. One of the major advantages of VCSELs is that they can be fabricated into arrays, and locking such arrays will be a major focus of the project. This would enable a 1 Tb/s WDM transmitter on a chip, and higher speed optical interconnects, for example. Arrays will be fabricated with the lasers fiber coupled to a multiplexer (AWG), and injection-locked using an optical comb generator (such as a mode-locked laser). We will integrate fibers bundles with the VCSEL arrays by using novel packaging techniques. Further, the arrays can also be integrated with CMOS circuitry for digital optical interconnect applications, as well as with very high frequency RF CMOS for 0-50 GHz radio-over-fiber applications, ex. wireless signal delivery.
The development of these transmitters will enable new applications and network designs. Novel networking experiments involving techniques such as coherent communications and all-optical optical clock recovery will be performed using the devices developed in this program. Finally, the large scale arrayed transmitters will be tested and incorporated in networking demonstrations to create much higher bit-rate systems, and ultimately will find their way into commercial communication networks.
Reference to our work in Laser Focus World