Antenna model patterns improvement through Error Matrix refinement

Abstract:
Accurate knowledge of antenna patterns is of main importance for SAR data processing, in order to have the possibility to extract unbiased geophysical information from the acquisitions. Traditional antenna pattern monitoring is performed by estimates obtained from real data acquisitions over distributed areas such as rain forest. However, the rise of modern sensors enabling a large number of beam acquisitions and the need of short calibration duration, made necessary the development of advanced calibration techniques. In this work, we propose the use of a numerical optimizer able to refine the inputs of the antenna model in order to find a refined input for the traditional antenna model technique. Experiments on real data confirm the effectiveness of the proposed methodology.