A fast integral equation based method for solving electromagnetic inverse scattering problems with inhomogeneous background - CentraleSupélec Accéder directement au contenu
Article Dans Une Revue IEEE Transactions on Antennas and Propagation Année : 2018

A fast integral equation based method for solving electromagnetic inverse scattering problems with inhomogeneous background

Résumé

A family of difference integral equations, consisting of difference Lippmann-Schwinger integral equation (D-LSIE) and difference new integral equation (D-NIE), is proposed to solve the electromagnetic inverse scattering problems (ISPs) with inhomogeneous background medium bounded in a finite domain. Without resorting to Green’s function for inhomogeneous background medium, in the frame of the difference integral equation methods, the Green’s function with homogeneous medium is utilized such that not only fast algorithms (referring to those used in forward scattering problems, like CG-FFT, FMM) can be adopted but also the burdensome calculation for the numerical Green’s function for the inhomogeneous background medium is avoided. Especially, to tackle the ISPs with strong non-linearity, those with large contrast and/or large dimensions, a Low-Pass Filter-Matching (LPFM) regularization is introduced, which aims to stably match the information from the background medium to the unknown scatterer. Together with the D-NIE model, the proposed inversion method can efficiently tackle the ISPs with strong non-linearity while a bounded inhomogeneous medium being present. Against both synthetic and experimental data, several representative numerical tests illustrate the efficacy of the proposed inversion method.
Fichier non déposé

Dates et versions

hal-01634160 , version 1 (13-11-2017)

Identifiants

Citer

Kuiwen Xu, Yu Zhong, Xudong Chen, Dominique Lesselier. A fast integral equation based method for solving electromagnetic inverse scattering problems with inhomogeneous background. IEEE Transactions on Antennas and Propagation, 2018, 66 (8), pp.4228-4239. ⟨10.1109/TAP.2018.2839763⟩. ⟨hal-01634160⟩
264 Consultations
0 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More