No abstract

We report programmable shaping of femtosecond optical pulses by use of a multielement liquid-crystal modulator to manipulate the phases of spatially dispersed optical frequency components. Our approach provides for continuously variable control of the optical phase and permits the pulse shape to be reconfigured on a millisecond time scale. We use the apparatus to demonstrate femtosecond pulse-position modulation as well as programmable compression of chirped femtosecond pulses. Considerable effort has been directed toward generating ever shorter optical pulses, and pulses as short as 6 fsec are now available.' More recently, interest in synthesis of ultrashort pulses with arbitrarily controllable pulse shapes 2-6 has also arisen. Specially shaped ultrashort pulses are now being used to study highspeed optical communications, 4 nonlinear optics in fibers, 7-9 and time-resolved spectroscopy. 8 "l 0

Coupling of two-dimensional (2D) vertical-cavity surface-emitting lasers (VCSELs) to give a coherent supermode is described. The top metal layer of a stained-layer InGaAs quantum well VCSEL structure was patterned laterally by depositing various metals with different optical reflectivities. This lateral reflectivity patterning defined a 2D laser array sharing the same ‘‘supercavity’’. It is shown that these 2D arrays oscillate in a stable single, coherent 2D supermode. This was achieved with a simple planar process and without significant deterioration of threshold current and efficiency relative to an equivalent broad-area VCSEL.

An optical learning machine that has multicategory classification capability is demonstrated. The system exactly implements the single-layer perceptron algorithm and is fully parallel and analog. Experimental results on the learning by examples obtained from the system are described.

Previous research assessing planar holographic storage and cross talk between volume multiplexed holograms is reviewed. Using these results, we derive equations for volume holographic storage density and capacity. These equations permit us to identify and to calculate the trade-off between storage density and input-output parallelism and the dependence of capacity on both the parallelism and the resolution of the input-output path.