Super-resolution spectral estimation methods for buried and through-the-wall object detection

Abstract

Seeing the targets behind and inside visually opaque obstacles such as walls using microwave signals is considered as a powerful tool for a variety of applications in both military and commercial paradigms. The ultimate goal in Through-the-Wall Object Detection (TWOD) and Buried Object Detection (BOD) systems is to achieve High Range Resolution (HRR). HRR provides the ability of resolving closely spaced targets in range, improves the accuracy of range estimates and aids in target recognition and classification. HRR can be achieved using impulsive waveforms which use extremely narrow pulses, frequency modulated waveforms which increase the instantaneous bandwidth by applying frequency modulation to each transmitted pulse, stepped-frequency waveform, and signal processing techniques. The range resolution of stepped-frequency and frequency modulated continuous wave (FMCW) radar systems is limited by the Inverse Fast Fourier Transform (IFFT) and Fast Fourier Transform (FFT), respectively. FFT provides poor range resolution for data with a small bandwidth and when the data size is small. On the other hand, it is well known that parametric spectral estimation methods provide super-resolved range profiles of the targets compared with FFT for the same frequency bandwidth. This thesis studies the target detection and range extraction performance of ESPRIT, Root-MUSIC, Higher Order Yule-Walker, Minimum-Norm, Yule-Walker, and Least- Squares methods in BOD and TWOD applications using synthetic stepped-frequency and FMCW radar signals and experimental stepped-frequency radar data.