| While the majority of modern experimentation in optics and optical technology relies on pure and highly coherent sources, the light encountered in nature is of inferior quality. The low-quality, or noisy, light creates problems in nonlinear signal processing, as the random, multi-mode distribution inhibits phase matching and wave mixing. In this dissertation, we will discover new incoherent phenomena using nonlinear optics and characterize many fundamental, and useful, features pertinent to waves with inferior coherence.The first part of the thesis will be devoted to a new theory describing the nonlinear propagation of statistical light. The essence of the theory is to represent incoherent light as a gas of particles (speckles) that can interact collectively via nonlinearity, effectively forming a photonic plasma. We carried out a set of basic plasma-like experiments in optics and showed that this representation is valid and promising. Experiments were conducted using a nonlinear photorefractive crystal and basic phenomena such as modulation and bump-on-tail instabilities, optical turbulence, etc., were observed.In the second part of the thesis, we will apply this plasma formalism to the recovery and amplification of weak, noise-hidden images. The signal fidelity will be shown to improve by exploiting signal-noise interaction in the nonlinear medium. This new, dynamical type of stochastic resonance (a process in which signal can grow at expense of the noise) is treated as an equivalent beam-plasma instability, allowing an analytical characterization of the resonance as a function of coupling strength, noise statistics, modal content of the signal and wavelength. The theory also suggests an exponential limit to the amount of information transmissible in nonlinear communications systems. The results link the fields of optics, plasma and information theory, and pave the way for a variety of nonlinear, instability-driven imaging techniques. |
