• DocumentCode
    3676396
  • Title

    Minimal-rank ℋ2-matrix-based iterative and direct volume integral equation solvers for large-scale scattering analysis

  • Author

    Dan Jiao;Saad Omar

  • Author_Institution
    School of Electrical and Computer Engineering, Purdue University, West Lafayette, IN 47907, USA
  • fYear
    2015
  • fDate
    7/1/2015 12:00:00 AM
  • Firstpage
    740
  • Lastpage
    741
  • Abstract
    It can be shown that the matrix structure resulting from a fast multipole method (FMM)-based algorithm is an ℋ2-matrix, but with a full-rank representation for electrically large analysis. We compare the computational complexity of a volume integral equation (VIE) solver having a minimal-rank ℋ2-representation with that of a VIE solver using an FMM-based ℋ2-representation. The former is shown to be strict O(N) in storage and matrix-vector multiplication, and O(NlogN) in inverse irrespective of electric size. Such a complexity has been demonstrated by the analysis of large-scale dielectric scattering problems involving millions of unknowns.
  • Keywords
    "Complexity theory","Accuracy","Dielectrics","Sparse matrices","Integral equations","Couplings","Scattering"
  • Publisher
    ieee
  • Conference_Titel
    Antennas and Propagation & USNC/URSI National Radio Science Meeting, 2015 IEEE International Symposium on
  • Type

    conf

  • DOI
    10.1109/APS.2015.7304757
  • Filename
    7304757
    • Link To Document
    https://search.isc.ac/dl/search/defaultta.aspx?DTC=49&DC=3676396