Theory and Application of Wavelet Based Bi-Orthonormal Decomposition Method in the Solution of Linear Inverse Problems in Electromagnetics

XJ ZHU,GW PAN
DOI: https://doi.org/10.1109/aps.1994.407793
1994-01-01
Abstract:A wavelet based bi-orthonormal decomposition method is proposed and applied to the solution of linear inverse problems in electromagnetics. The unknown function is expanded into wavelets, where an adaptive algorithm is developed utilizing the multiresolution properties of the wavelet. The spectral domain method is used to find the bi-orthonormal bases for shift-invariant operators; and the wavelet decomposition method is formulated to construct the biorthonormal bases for general operators. The modified least square QR iterative method is employed to solve the resulting sparse matrix equations. Finally the method is used in the computation of the scattering of TM waves by an elliptic conducting cylinder.<>
What problem does this paper attempt to address?