Comparison of SAR simulation concepts for the analysis of high-resolution SAR data

The simulation of SAR images provides valuable information used in many fields such as mission planning or the design of feature extraction algorithms. Simulation is in general a trade-off between achievement of appro-ximation to reality as close as possible and minimization of computational load respectively processing time. For real-time application, a fast but suboptimal approach might be optimal, whereas for thorough parameter tuning of image processing or image analysis algorithms, more sophisticated methods would be required. In this paper, three SAR simulators using different simulation concepts tailored for a variety of tasks are introduced and the simulation results for simple building models and complex urban scenes are compared. To evaluate the realism of the simulation methods, the results are compared to a real SAR image of an urban scene. On this basis the possibilities and limitations of the different SAR simulation concepts are discussed.