Fast computation of angular responses of large-scale three-dimensional electromagnetic wave scattering

Zhen Peng, Mattew B. Stephanson, Jin Fa Lee

Research output: Contribution to journalArticlepeer-review


Many applications in electromagnetic engineering require predicting the currents induced by a wide range of incident plane waves on a given object. This paper proposes a way to efficiently and accurately compute the angular responses for electrically large targets. Since the targets of concern are electrically large, we have employed the newly developed non-conformal domain decomposition method (DDM) as the direct Maxwell equation solver. The proposed fast angular sweep algorithm is then applied in conjunction with the DDM solver to produce wide range of angular responses. There are basically three main ingredients in the proposed fast angular sweep algorithm: the use of Fourier harmonics to interpolate the right-hand-sides (RHSs) over a given angular range; the number of independent coefficient vectors in the Fourier harmonics can be further reduced through the application of singular-value-decomposition (SVD); and, finally, the matrix solution of multiple RHSs can be solved efficiently using a recently proposed Krylov recycling method, GCRO-DR. Three numerical results are presented and compared to measurements and direct computations to validate the numerical solutions and to illustrate the effectiveness of the proposed approach.

Original languageEnglish (US)
Article number5484634
Pages (from-to)3004-3012
Number of pages9
JournalIEEE Transactions on Antennas and Propagation
Issue number9
StatePublished - Sep 2010
Externally publishedYes


  • Angular response
  • Maxwell's equations
  • domain decomposition
  • electromagneitc wave scattering
  • recycling Krylov space method

ASJC Scopus subject areas

  • Electrical and Electronic Engineering


Dive into the research topics of 'Fast computation of angular responses of large-scale three-dimensional electromagnetic wave scattering'. Together they form a unique fingerprint.

Cite this