Coherence in Dynamic Metasurface Aperture Microwave Imaging Systems

Abstract

Microwave imaging systems often utilize electrically large arrays for remote characterization of spatial and spectral content. Image reconstruction involves computational processing, the success of which depends on adequate spatial and temporal sampling at the array as dictated by the nature of the radiation and sensing strategy. Effective design of an imaging system, consisting of its hardware and algorithmic components, thus requires detailed understanding of the nature of the involved fields and their impact on the processing capabilities. One can sufficiently characterize many of the properties of such fields and systems via coherence, which quantifies their interferometric capacity in terms of statistical correlations and point spread functions. Recent work on dynamic metasurface apertures (DMAs) for microwave imaging has demonstrated the utility of these structures in active, coherent systems, supplanting traditional array architectures with lower-cost designs capable of powerful wavefront shaping. In contrast to arrays of distinct antennas, DMAs are composed of electrically large arrays of dynamically-tunable, radiating metamaterial elements to realize diverse gain patterns that can function as an encoding mechanism for coded aperture image reconstruction. With the appropriate formulation, well-established concepts from the realm of Fourier optics can be transposed from conventional array systems to DMA architectures. This dissertation furthers that task by modeling DMA imaging systems involving partial coherence and incoherence, and demonstrating new reconstruction algorithms in these contexts. Such an undertaking provides convenient opportunities for examining the origins of coherence in computational and holographic imaging systems. This insight is necessary for the development of modern approaches that seek to smoothly integrate hardware and computational elements for powerful, efficient, and innovative imaging tasks. Accommodating different degrees of coherence in a microwave imaging system can substantially relax demands on hardware components including phase stability and synchronization, or on algorithmic procedures such as calibration. In addition, incoherent operation can yield improved images free of coherent diffraction artifacts and speckle. Finally, an understanding of coherence can unlock fundamentally distinct applications, such as passive imaging and imaging with ambient illumination, that can benefit from the flexibility of a DMA system but have yet to be demonstrated under such an architecture. To this end, I formulate a unified framework for analyzing and processing array imaging systems in the Fourier domain, and demonstrate a method for transforming a DMA-based system to an equivalent array representation under active, coherent operation. I then investigate the role of spatial coherence in a two-dimensional holographic imaging system, and experimentally demonstrate some results using a collection of DMAs. I conduct a similar investigation in the context of single-pixel ghost imaging, which allows coherent and incoherent imaging directly from intensity measurements, thereby relaxing hardware phase requirements. I then formulate a model for partially coherent fields in a DMA imaging system, and provide several reconstruction strategies and example simulations. I finally restrict this general case to passive, spectral imaging of spatially and temporally incoherent sources, and experimentally demonstrate a compressive imaging strategy in this context.