Efficient Azimuth Image-Quality Prediction for Block-Design in Squinted Spotlight SAR Imaging

This paper presents an efficient image-quality prediction method for planning block processing in squinted spotlight synthetic aperture radar (SAR) imaging. Before real SAR data are divided into processing blocks, candidate processing-reference locations and feasible block extents must be evaluated. Repeated evaluation using a full two-dimensional focusing processor is computationally expensive and is therefore inefficient for iterative planning. To overcome this problem, first, a new one-dimensional azimuth-compression kernel is derived by applying the principle of stationary phase (POSP) to the azimuth-phase component of the SAR signal. Second, the possibility of replacing the full two-dimensional ω–k kernel with the derived kernel is mathematically analyzed by examining the residual quadratic phase and its range-frequency dependence. Third, applying the derived kernel to initial block design substantially reduces the computation time. In the present study, idealized platform trajectory and ground-surface conditions are assumed. Under these assumptions, the proposed kernel predicts azimuth resolution, peak sidelobe ratio (PSLR), and integrated sidelobe ratio (ISLR). These metrics are then used as image-quality criteria for initial block design. Simulations using 36 point targets distributed over a 1.5 km × 1.5 km test area demonstrate that the proposed 1D kernel closely matches the image quality obtained using the full ω–k kernel. These results demonstrate that the proposed 1D azimuth-compression kernel can efficiently replace full ω–k processing during initial block design.