A COMPACT OPTICAL CROSSBAR SWITCH ARRAY FOR LARGE-SCALE PHOTONIC NETWORKS

THE LARGE AMOUNT OF DATA GENERATED IN HIGH ENERGY PHYSICS EXPERIMENTS NECESSITATES DATA PROCESSING TECHNIQUES CAPABLE OF HANDLING VERY HIGH DATA RATES. OPTICAL TECHNIQUES FOR DATA TRANSMISSION AND FOR NETWORKING ARE DESIRABLE BECAUSE OF THEIR LARGE BANDWIDTH, HIGH SPEED, LOW LOSS, AND RESISTANCE TO ELECTROMAGNETIC NOISE. ONE OF THE DRAWBACKS OF OPTICAL NETWORKS IS THE LACK OF LARGE-SCALE PHOTONIC SWITCHING ARRAYS, ANALOGOUS TO THE VERY LARGE-SCALE INTEGRATED CIRCUIT CROSSBARS IN ELECTRONIC PROCESSING. MANY OPTICAL CROSSBAR SWITCH TECHNIQUES PASSIVELY SPLIT AND COMBINE THE OPTICAL BEAMS, WITH ACTIVE MODULATION PERFORMED BY A SPATIAL LIGHT MODULATOR. THESE SWITCHES HAVE SIGNIFICANT POWER BUDGET PROBLEMS BECAUSE OF THE FAN-OUT AND FAN-IN LOSS, AND THEREFORE CAN HANDLE ONLY A LIMITED NUMBER OF CHANNELS. FOR THIS REASON, OPTICAL FIBER NETWORKS SHOULD HAVE NONBLOCKING SWITCHES THAT TRANSFER THE ENTIRE SIGNAL INTO THE SELECTED CHANNEL. THIS PROJECT PLANS TO DEVELOP A LARGE-SCALE PHOTONIC SWITCHING ARRAY FOR HIGH DATA RATE FIBER OPTIC NETWORKS OR COMPUTER INTERCONNECTS. THE SWITCH IS BASED ON THE PRINCIPLE OF MATRIX VECTOR MULTIPLICATION. IN THE PLANNED APPROACH, HOWEVER, THE FUNCTION OF BEAM SPLITTING AND COMBINING IS EMBEDDED IN THE FUNCTION OF THE SWITCH ITSELF. THIS WILL ELIMINATE THE FAN-IN AND FAN-OUT LOSS AND PERMIT THE DESIGN OF LARGE-SCALE PHOTONIC SWITCHING ARRAYS. IN PHASE I, THE DESIGN OF THE SWITCH IS BEING PRODUCED AND PRELIMINARY EXPERIMENTS WILL BE PERFORMED TO DEMONSTRATE THE FEASIBILITY OF THE SWITCH. A PROTOTYPE DEVELOPMENT IS SCHEDULED FOR PHASE II.