Generalized filtering configurations with applications in digital and optical signal and image processing

In this thesis, we first give a brief summary of the fractional Fourier transform which is the generalization of the ordinary Fourier transform, discuss its importance in optical and digital signal processing and its relation to time-frequency representations. We then introduce the concept of filtering circuits in fractional Fourier domains. This concept unifies the multi-stage (repeated) and multi-channel (parallel) filtering configurations which are in turn generalizations of single domain filtering in fractional Fourier domains. We show that these filtering configurations allow a cost-accuracy tradeoff by adjusting the number of stages or channels. We then consider the application of these configurations to three important problems, namely system synthesis, signal synthesis, and signal recovery, in optical and digital signal processing. In the system and signal synthesis problems, we try to synthesize a desired system characterized by its kernel, or a desired signal characterized by its second order statistics by using fractional Fourier domain filtering circuits. In the signal recovery problem, we try to recover or estimate a desired signal from its degraded version. In all of the examples we give, significant improvements in performance are obtained with respect to single domain filtering methods with only modest increases in optical or digital implementation costs. Similarly, when the proposed method is compared with the direct implementation of general linear systems, we see that significant computational savings are obtained with acceptable decreases in performance.