Effect of instrument frequency uncertainty on wideband microwave synthetic aperture radar (SAR) images

Abstract

In this paper, we investigate the effect of frequency uncertainty in signals generated or measured by a microwave instrument on the resulting synthetic aperture radar (SAR) images, particularly for nondestructive testing (NDT) applications. Wideband SAR imaging systems measure reflections from a target by irradiating it with locally generated signals that can potentially have some level of frequency uncertainty. Quantifying this frequency uncertainty provides the user with a realistic and expected level of image distortion which may manifest itself as blurring, noise artifacts, etc. In this paper, we show that as uncertainty in the actual frequency value increases, the level of image distortions increases predominantly for distant targets. This is an important fact for NDT applications since the imaged object is commonly close to the imaging system. In addition, these imaging system usually have a limited “aperture” size, which makes target distance an important consideration. For complex targets, we show how frequency uncertainty-based image distortions can dominate features in an image depending on the reflected signal amplitude from the target. We also show that in real imaging systems, the statistical distribution of frequency uncertainty combined with practical, near-target ranges (distances) produce nondiscernible image distortions.