Synthetic Aperture Radar Processing with GPGPU

Synthetic aperture radar (SAR) processing is a complex task that involves advanced signal processing techniques and intense computational effort. While the first issue has now reached a mature stage, the question of how to produce accurately focused images in real time, without mainframe facilities, is still under debate. The recent introduction of general-purpose graphics processing units (GPGPUs) seems to be quite promising in this view, especially for the decreased per-core cost barrier and for the affordable programming complexity. This article focuses on methodologies with recurrent use to code examples that try to couple with the flow of the main steps of the SAR processing. The possibility to be comprehensive was prevented by the wide scenario of variations of the focusing algorithm as well as the spread of applications. The reader should look at this work as a sample of possibilities offered by this new technology and a collection of suggestions and considerations that may guide to new applications and horizons.