• DocumentCode
    2146864
  • Title

    Parallel Electromagnetic Simulation for Electric Large Target by EMS-FMM

  • Author

    Wang, Wu ; Feng, Yangde ; Chi, Xuebin

  • Author_Institution
    Supercomput. Center, CAS, Beijing, China
  • fYear
    2010
  • fDate
    18-22 Aug. 2010
  • Firstpage
    36
  • Lastpage
    39
  • Abstract
    The development framework of parallel electromagnetic simulation software EMS-FMM based on multilevel fast multipole method (MLFMM) is present in this paper. EMS-FMM can solve large-scale scattering problem of 10 million scale with a complexity of O (Nlog N), so it´s efficient for electric large target. It´s implemented by C and Fortran with MPI communication on cluster DeepComp 7000. Implementation of MLFMM, interpolation algorithm for sphere functions, load balance and communication scheme for distributed octree structure are also discussed. The results demonstrate that surface current and radar cross section (RCS) of large complex target, such as plane and destroyer models, can be calculated by EMS-FMM.
  • Keywords
    C language; application program interfaces; interpolation; octrees; parallel programming; radar cross-sections; resource allocation; scattering; C; EMS-FMM; Fortran; MPI communication; cluster DeepComp 7000; distributed octree structure; electric large target; interpolation algorithm; large-scale scattering problem; multilevel fast multipole method; parallel electromagnetic simulation; radar cross section; sphere functions; surface current; Clustering algorithms; Complexity theory; Interpolation; Mathematical model; Octrees; Program processors; EMS-FMM; electric large target; multilevel fast multipole method; parallel implementation;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Frontier of Computer Science and Technology (FCST), 2010 Fifth International Conference on
  • Conference_Location
    Changchun, Jilin Province
  • Print_ISBN
    978-1-4244-7779-1
  • Type

    conf

  • DOI
    10.1109/FCST.2010.43
  • Filename
    5576127