Multipath Model Enhanced SBR Technique for Prediction of Near-Field EM Scattering From Objects on Rough Surfaces

Both the four-path-based multipath signal model and existent shooting and bouncing ray (SBR) techniques are widely used in electromagnetic (EM) scattering signature prediction for electrically large objects on rough surfaces. Either has its limitation, i.e., the lack of higher order interactions in the multipath model and the burdensome or even impractical ray tracing in SBR for low grazing angle cases where extremely large scenes must be considered. In this article, a multipath model enhanced SBR (MP-SBR) technique is proposed to deal with near-field EM scattering prediction for an object on rough surface in both low and high grazing angle cases. A modified rough surface reflection coefficient calculation formulation, which is employed to statistically quantify the coherently reflected fields from the rough surface, is first derived to consider the near-field effect on the multipath scattering fields. Then, the multipath model with the modified reflection coefficient is inserted into the near-field SBR process for a more exact while still efficient calculation of the scattered fields from an object on rough surface, including those not only stimulated from the direct incidence but also from the surface reflected illumination, which originates from the rough surface within the far-away Fresnel zone, as well as the EM interactions among the object and its nearby areas of the underlying surface. The results calculated using the method of moments (MoM) are used to validate the MP-SBR technique. Simulated radar images are also presented and compared with the measured counterparts to demonstrate the usefulness of the proposed technique in high-fidelity radar image generation.