RCS interpolation in frequency and angle using adaptive feature extraction

Abstract

We address the interpolation problem in frequency and angle using a model-based approach. Our approach is the adaptive feature extraction (AFE) algorithm. It has been used previously by us to eliminate the aliasing effect and construct ISAR image from unevenly undersampled measurement data. Unlike standard interpolation algorithms which suffers from the Nyquist sampling limitation, the AFE can overcome the Nyquist sampling criterion by using uneven sampling. The essential idea of the adaptive algorithm is to search and extract out individual scattering features one at a time in an iterative fashion. When applied to the present problem, the interference between different scattering features can be avoided. After all the main features are extracted, the current and the RCS on a denser grid of sampling can be interpolated by summing the contributions from all the extracted scattering features. The AFE algorithm is tested using numerical examples for both 1-D frequency interpolation and 2-D frequency-aspect interpolation.