Microlocal ISAR for low signal-to-noise environments

We consider the Inverse Synthetic Aperture Radar problem in which high-range-resolution radar pulses interrogate a rotating target, and the measured echos are then used to isolate target features. Our approach to this problem is based on the techniques of microlocal analysis, which is a mathematical theory developed to handle high-frequency asymptotics. In essence, our approach [6], [7] is to relate high-frequency components of the data to features of the target. Because this scheme applies directly in the data domain, it may have a number of advantages over conventional imaging methods. The purpose of this paper is to show how this theory can be applied to realistic band-timited data. In particular, we propose an iterative algorithm (based on the generalized Radon-Hough transform) in which we estimate the target features associated with high-frequency components of the data, one after another, and subtract out the corresponding band-limited components. We show the results of numerical tests on simulated noisy data.