Phase Errors and Autofocus in SAR Imagery

Abstract

In previous chapters, we discovered how the steps involved in the collection and formation of SAR imagery can be explained in light of tomographic principles. We also saw that several assumptions were made in order for that paradigm to be valid. One of these assumptions limited the allowable amount of wavefront curvature; a second assumption required that the residual phase term from the deramp processing be sufficiently small. We demonstrated how the reconstructed image quality is degraded when these assumptions are not met and derived quantitative conditions on patch size, resolution, standoff range, and several other parameters (see Equations 2.74 and 2.75) that render the degradation in image quality negligible. A third assumption dealt with robust image formation. This assumption is that the amount of time required for each radar pulse to travel from the SAR platform to the patch center and back is known precisely for each transmission point along the synthetic aperture. This demodulation time was defined as τ₀ = 2R ₀ /c, where R ₀ is the distance from the SAR platform to the patch center and c is the propagation velocity of the electromagnetic wave. Up to this point, we simply ignored the consequences of inaccuracies in demodulation times that can arise from noise in real SAR systems. In this chapter we address the matter in detail, because it becomes an important issue in the design and operation of real spotlight-mode SARs.