Multiple-input-multiple-output (MIMO) radar image processing presents problems difficult to address by modifying conventional monostatic radar methods as Fourier range migration. When the distance between the transmitter and receiver is comparable to the target size, the single phase center approximation is not accurate. Furthermore, if the antenna radiation pattern significantly deviates from a spherical wave, the symmetries assumed in most range migration techniques are violated. We present a rapid Fourier-based MIMO reconstruction called Fourier accelerated multistatic imaging (FAMI) suitable for massively parallel computation that accounts for frequency-dependent radiation patterns, does not require the single phase center approximation, and is able to dynamically adapt to different target support volume shapes. FAMI is especially suitable for frequency-diversity antenna systems that use spectrally modulated coded spatial radiation patterns.
- GPU acceleration
- GPU computing