Efficient hybrid heterogeneous parallel computing method for electromagnetic scattering and diffraction of electrically large objects

The main mechanisms of high-frequency electromagnetic scattering include specular reflection, edge diffraction, and creeping wave diffraction, among others. To overcome the limitations of single-algorithm methods for electrically large targets, which suffer from incomplete mechanisms and an inherent trade-off between efficiency and accuracy, this paper proposes a collaborative modeling framework with heterogeneous parallel acceleration. The framework integrates the multilevel fast physical optics (MLFPO) method for specular scattering, the truncated wedge incremental length diffraction coefficient (TWILDC) method for edge diffraction, and the incremental length diffraction coefficient (ILDC) method for creeping waves, establishing a comprehensive physical model that enhances accuracy. For efficiency, a fine-grained GPU parallelization strategy based on the CUDA and an event-triggered synchronization mechanism enable effective CPU-GPU co-processing. Numerical examples for electrically large targets demonstrate that the proposed collaborative model achieves excellent agreement with the multilevel fast multipole algorithm (MLFMA), reducing the mean absolute error by 2.98dB versus the standalone MLFPO. The proposed heterogeneous parallel scheme delivers a maximum 318.43× speedup over MLFMA, exceeding two orders of magnitude. This work provides robust support for high-confidence radar cross section (RCS) prediction of complex electrically large targets.