M. Shabeer

Interferometric noise reduction through intrabit frequency evolution of directly modulated DFB lasers

A new technique is demonstrated that improves the performance of optical networks subject to interferometric noise. Every transmitter comprises a selected DFB laser that is NRZ ASK directly modulated with a large modulation depth. Under modulation, the center frequency of the DFB laser is found to vary over the duration of each bit, falling during mark bits and rising during space bits with an exponential-like dependency of characteristic time constant /spl sim/20 ns. The optical frequency at the midpoint of each bit interval is not constant but depends upon the sequence of the preceding bits. Therefore, on the interference of a delayed-replica parasitic crosstalk waveform the interferometric noise generated on detection may be reduced by RF filtering according to the difference in center frequency of the interfering bits. This noise suppression which requires no additional hardware has been successfully modeled and experiment demonstrates little performance gain for a single interferer, as predicted. However, in the presence of multiple interferers significant improvement is predicted and observed at sub-Gb/s rates.

Sequence dependence of phase-induced intensity noise in optical networks that employ direct modulation.

Phase-induced intensity noise in optical networks that employ directly modulated laser sources is observed to be bit-sequence dependent. This dependence is explained by optical frequency variations that are due to the heating history of the laser chip and is accurately modeled. This effect may permit suppression of phase-induced intensity noise in many types of fiber system with multipaths.

Fiber-optic bipolar tap implementation using an incoherent optical source.

A fiber-optic bipolar tap implementation is described. The proposed scheme utilizes 2s complement-number representation and digital multiplication by an analog convolution algorithm. Simple experimental verification of the scheme is presented. The addition of programmability to the taps results in the realization of a fast, flexible processor for manipulating binary quantities.

Fiber-optic matrix multiplier using a two-dimensional systolic-array architecture.

An optical matrix multiplier based on a two-dimensional systolic-array architecture is described, exploiting the parallelism offered by optical methods. Experimental verification of multiplier performance is presented. A technique for multiple matrix multiplications is suggested.

Modeling the performance of optical time division multiplexed (TDM) switching structures in the presence of interferometric noise due to interchannel crosstalk

Crosstalk induced interferometric noise has been found to constrain the performance of optical switching networks. A computer design and simulation package has been used to analyze the effect of this noise on optical TDM crossconnects. This software tool is able to calculate all possible routes that a signal may traverse within a crossconnect for a given input/output assignment, and also incorporates a theoretical analysis of interferometric noise. It has been found that for a crossconnect capable of switching 512 independent channels, the constituent 2 x 2 crosspoints must possess isolations better than -20 dB, and the switch architecture must be dilated.

All optical digital correlators for multiple access networks

An all-optical guided-wave approach in which a network is implemented and programmed to operate upon a data stream in the link and will determine a match or mismatch condition is described. The basic building block consists of two conventional guided-wave directional couplers interconnected with a predetermined time delay T, corresponding to the delay between incoming optical bits. Each coupler is programmed in the following manner: for a coupler to represent 1 the coupling ratio is set, for example, to 5% and for a binary 0 the coupler is set to 0% coupling. The coupled channels are interconnected by a length of waveguide, the time delay of which is smaller than T, and are fed into a thresholding device which acts as a limiter. The techniques employed are incoherent, obviating the severe operational problems inherent in coherent approaches.<<ETX>>

Fibre Optic Signal Processing

Fibre optic technology lends yet another dimensions to the topic of optical signal processing complementing both Fourier plane and digital optical approaches. It is inevitable that eventually optical signal processing will combine all three of these basic concepts. However, in this paper we consider the possibilities offered by the fibre optic transmission medium in the context of signal processing. The unique features of the transmission medium imply that precise frequency domain filters may be readily realised. Furthermore, functions such as matrix-matrix multiplication and code recognition may also be programmed in the all optical fibre form. These processes are essentially linear but the addition of a simple non-linear element may yet further extend the capability of these basic concepts. This paper outlines the general principles underlying fibre optic signal processing and gives some indication of its performance potential.

Programmable fiber/integrated-optics bipolar tap

A programmable bipolar tap implementation is described. The proposed scheme utilizes fiber and integrated-optics technology in conjunction with twos complement number representation and digital multiplication by an analog convolution algorithm. Simple experimental verification of the scheme is presented.