Spectral Estimation for Synthetic Aperture Radar

Date of Award

5-9-2026

Degree Name

M.S. in Electrical Engineering

Department

Department of Electrical and Computer Engineering

Advisor/Chair

Brian Rigling

Abstract

Synthetic aperture radar (SAR) image formation estimates spatial frequency content from finite, noisy measurements. This dissertation reformulates SAR image formation across multiple imaging modalities as a spectral estimation problem, enabling the application of modern adaptive spectral estimation techniques to improve image quality. A unified framework is developed to implement these methods and analyze their impact on image formation and downstream processing tasks, including interferometry, tomography, and automatic target recognition. Emphasis is placed on computationally and memory-efficient implementations to enable practical deployment in real-world systems. A model for monostatic SAR image formation is introduced, and, based on this model, the image formation procedure is framed as a spectral estimation problem. Conventional imaging algorithms used in an image formation processor are shown to correspond to classical Fourier spectral estimation methods, which suffer from high variance, limited resolution, and poor sidelobe behavior due to a finite aperture. Although modern spectral estimation techniques provide improved statistical stability, conventional formulations typically discard phase information, limiting their applicability to coherent SAR processing. To address this limitation, spectral estimation methods are developed that retain complex phase information while improving image quality for both single- and multi-channel imaging geometries. A hybrid spectral estimation approach is introduced for stabilizing complex SAR imagery in the single-channel imaging mode. Then, two novel algorithms are introduced for dual-channel interferometric processing with application to both height estimation and ground moving target indication (GMTI). A multi-channel extension for the algorithms is then presented to further improve GMTI performance. The interferometric processing methods are extended to coherent change detection by reinterpreting the coherence image as a coherence spectrum estimation problem. The single channel imaging framework is further extended to three-dimensional tomographic SAR imaging. Finally, the impact of spectral estimation–based image formation on ATR performance is analyzed.

Keywords

Electrical Engineering

Comments

OCLC No. 1591628133

Rights Statement

Copyright 2026, author.

Link to Full Text

Share

COinS
 
 
 

Links