Iterative Physical Optics And Its Acceleration Methods In Electromagnetic Scattering From Target And Rough Sea Surface

Electromagnetic scattering from electrically large rough surface and targets has been extensively investigated. Solutions to electrically large scattering problems are time-consuming, and the time of occlusion judgment is also considerable. Parallel computing has recently been applied in many fields with the improvement of computer hardware. For the iterative physical optics(IPO) method, the update of electromagnetic current on the surface of an object is a traversal and time-consuming process. Therefore, the corresponding accelerated technique is employed to improve the computational efficiency in this paper, which is based on the Open GL and parallel technology.Firstly, the theory of Gaussian rough surfaces, sea spectra and method of sea surface construction are introduced in this paper. Discrete points on the sea surface are transformed to a sea surface of grids, which are then transformed to a sea surface of NURBS in Rhinoceros5.0 with the use of Rhino Resurf. The composite model is achieved by combining the sea and the target together. In addition, traditional occlusion judgment is also described, and the processes of the algorithms are given. Then, the principle of physical optics is introduced, and several examples of electromagnetic scattering of objects by this method are presented. However, precision of the result is not good when coupling effect exists. Thus, IPO which considers coupling effect is used to improve the precision.However, the occlusion judgment should be performed before calculating the radar cross section of an object by IPO. Traditional occlusion judgment is time-consuming, so Open GL-based occlusion judgment is employed in this paper to accelerate the computing speed. For Open GL-based occlusion judgment, mesh data of objects are transferred into the Open GL program, and every mesh is printed with distinct color using the relation of the mesh number and color values. After orthographic projection, the pixel color values in projection window are read by functions in Open GL, and the mesh number are explored by the inverse operation of mesh number and color values. This process indicates the exploration of illuminated meshes of the object. Open GL-based occlusion judgment is more efficient than the traditional occlusion judgment.The speed of updating the electromagnetism of all meshes of an object can be accelerated by distributing the work into different threads using Open MP-based parallel technology based on shared memory technique. Moreover, the execution mode, compiler directive, API functions, and environment variables are also introduced in this paper. Finally, speed-up ratio, and parallel efficiency of the serial and parallel programs are presented. In conclusion, parallel program has the same accuracy as the serial program, but with decreased computing time. Therefore, the parallel IPO provides an effective way to solve the scattering problem of electrically large rough surface and targets.