Acousto-optical processors for Synthetic Aperture Radar (SAR) imaging are presented. The new processors produce images at real-time rates by combining the operations of data collection, storage, and processing into a compact time-and-space integrating (TSI) architecture. In the TSI approach the 2-D SAR imaging problem is decomposed into a cascade of 2 distinct operations: a 1-D spatial integration of light for range compression, and a 1-D temporal integration of light for azimuth compression. These two operations are coupled via a common path interferometric scheme that is insensitive to mechanical vibrations.
The results of an experimental characterization of the TSI approach, with simulated point scatterer radar echoes, are reported. The performance issues of interferometric bias removal, dynamic range, and resolution are addressed. The architecture is generalized to correct for range migration and the results of a range walk compensation experiment are presented. A programmable version of the TSI architecture, in which the fixed azimuth reference mask is replaced by an acousto-optic light modulator, is described. The application of the programmable architecture to both strip-map and spot-light mode SAR is analyzed and experimentally verified.
