Advances in Synthetic Aperture Radar from a Wavenumber Perspective

Abstract

This dissertation examines the wavenumber domain of Synthetic Aperture Radar (SAR) images. This domain is the inverse Fourier transform domain of a SAR image. The dissertation begins with the radar receiver's signal model and develops equations describing the wavenumber domain of a SAR image produced by a generalized bistatic and monostatic SAR system.

Then, closed form expressions for bistatic synthetic aperture radar spatial resolution of a generalized system from the wavenumber domain are developed. These spatial resolution equations have not previously appeared in the literature. From these equations, significant resolution is found in both range and cross-range forecasting a forward-scatter bistatic SAR image when the elevation angles of each bistatic platform are significantly different.

Next, wavenumber and time domain image formation algorithms are discussed. Developed within this dissertation is a wavenumber preprocessing method that increases the speed of the Back Projection Algorithm (BPA). This preprocessing method takes advantage of deramped SAR radar returns and their polar wavenumber format. This new algorithm is called the Fast Decimated Wavenumber Back Projection Algorithm (FDWBPA). Matlab functions are included to implement this algorithm, simulate bistatic SAR images and process the data from anechoic chamber tests demonstrating forward scatter resolution.