Acceleration Of High-frequency Electromagnetic Scattering Method Of Computing Research And Systems

RCS (Radar Cross Section) is a key factor in designing stealth weapons, so the research of radar targets scattering characteristics has become the foundation of stealth technologies. To get radar targets scattering information, computer simulation has been widely used nowadays, and the high frequency scattering predication methods have been proved effective in large complex targets electromagnetic scattering computation.Based on electromagnetic scattering theory, this paper mainly focuses on two principal scattering mechanisms, including facet scattering and multi-scattering. PO (Physical Optics) method is applied to calculate facet scattering and GO/PO (Geometrical Optics and Physical Optics) method for multi-reflection between facets. The above two parts are added together as the total electromagnetic scattering result. As the computation is intensified while the targets become larger and more complex, the original serial methods are ineffective in computationally demanding tasks. In this paper, we analyze and implement acceleration technologies for electromagnetic scattering computation.In theory, parallel computation could be implemented in space and time. Parallel in space means computation tasks are physically distributed with multiprocessors, while parallel in time means pipeline technology, which has been widely used in the field of computer graphics. Contemporary GPU (Graphics Processing Unit) has the parallel characteristics both in space and time, which is becoming a compelling platform in a wide variety of application domains. In this paper, we first analyze the parallelism in high frequency scattering computation, then we design the parallel methods based on GPU and PC-cluster, and finally compare the parallel results with those original serial ones. We validated that GPU and cluster are both effective ways to accelerate the electromagnetic scattering computation, which have been proved to have a good prospect.Applying the electromagnetic scattering theory with computing and visualization technologies, we developed an integrated high frequency scattering predication system. Our system is proved to be easier to reuse and maintain, as we use object-oriented methods to design the system.