Some issues in inverse synthetic aperture radar image reconstruction

One-dimensional radar target recognition schemes are based upon so-called `range profiles' and contain little or no cross-range target information. This limitation often restricts the usefulness of range-only methods in problems of target identification and target detection-in-clutter. Consequently, cross-range algorithms are usually required for these more demanding tasks. Because the ratio of wavelength to physical aperture is so much greater than it is for optical instruments of comparable dimensions, radar `imaging' methods usually attempt to artificially increase the effective aperture of the data acquisition system by employing SAR or ISAR techniques. Unfortunately, there are a host of problems associated with practical implementations of these SAR/ISAR techniques and the quality of the reconstructed image - as well as our ability to interpret it - can be seriously compromised. We present a brief overview of the ISAR techniques relevant to the problem of target recognition and pay special attention to scattering model limitations, sources of reconstruction error, and some current efforts to address these problems.