A Fast Algorithm for Solving the Inverse Scattering Problems with Inhomogeneous Background

Kuiwen Xu, Xiuzhu Ye, Zhong Yu, Xudong Chen

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

1 Citation (Scopus)

Abstract

In this paper, a family of difference integral equations, i.e., difference Lippmann-Schwinger integral equation (D-LSIE), is proposed to solve the the electromagnetic inverse scattering problems (ISPs) with inhomogeneous background medium. In the frame of the differential integral equation methods, the D-LSIE equipped with distorted-Born iterative method (DBIM) and the truanted singular value decomposition (TSVD), is used to solve the inhomogeneous ISPs. With the regularization of TSVD, Tikhonov regularization is not included in the cost function such that the method proposed avoid to choose a proper Tikhonov regularization parameter. The synthetic data illustrate the efficacy of the new inversion method.

Original languageEnglish
Title of host publication2018 IEEE International Conference on Computational Electromagnetics, ICCEM 2018
PublisherInstitute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)9781538612415
DOIs
Publication statusPublished - 17 Oct 2018
Externally publishedYes
Event2018 IEEE International Conference on Computational Electromagnetics, ICCEM 2018 - Chengdu, China
Duration: 26 Mar 201828 Mar 2018

Publication series

Name2018 IEEE International Conference on Computational Electromagnetics, ICCEM 2018

Conference

Conference2018 IEEE International Conference on Computational Electromagnetics, ICCEM 2018
Country/TerritoryChina
CityChengdu
Period26/03/1828/03/18

Keywords

  • Inverse scattering problems (ISPs)
  • difference integral equations
  • distorted-Born iterative method (DBIM)
  • inhomogeneous background
  • truanted singular value decomposition (TSVD)

Fingerprint

Dive into the research topics of 'A Fast Algorithm for Solving the Inverse Scattering Problems with Inhomogeneous Background'. Together they form a unique fingerprint.

Cite this