Synthetic aperture radar (SAR) imagery is formed using radar data collected from a moving platform (aircraft, vehicle, human, etc.). The radar transmits and receives backscatter signals in the down-range direction at a fixed pulse repetition interval (PRI) while the platform moves along the cross-range direction (called along-track) to generate a synthetic aperture. In the ideal situation, the platform moves at a constant speed and as a result, the radar will collect the phase-history data that are uniformly sampled along the aperture. However, in many situations the radar platform cannot be kept at a constant speed, e.g. a helicopter maneuvering over an imaging area for surveillance. The problem is even worse in the case of urban warfare with human-borne radar. A soldier moves at his own speed and creates erratic aperture sections with phase-history data that are either sparse or dense. The collected SAR data in such situation will result in SAR imagery with severe artifacts that might prevent us from detecting targets of interest.
In this paper, we will present the SAR imagery of non-uniform aperture data formed using the backprojection image formation algorithm. Although the backprojection image former is well suited to an arbitrary radar aperture, the SAR image artifacts are obvious from the nonuniform aperture. Using the nonuniform aperture phase-history data, we interpolate the data using a uniform grid along the aperture. We will show the resulting imagery with reduced artifacts. We use both simulated data and the Army Research Lab BoomSAR data to illustrate the artifacts generated by nonuniform sampling and the improvement using this interpolation technique.