Solutions of Microwave Problems Using VIE

Abstract

This thesis is an attempt to solve non trivial inverse problems using a global minimization technique as well as radial basis neural network. In general, we have two kinds of electromagnetic problems: forward problems and inverse problems. Forward problems are a calculation of scattered electromagnetic field when we know the material properties of the scatterer, The inverse problem is to find the material properties of the scatterer when we measure the scattered electric field due to a known electric field source. For the inverse problem solution, we need, at first, a forward problem solver. We use our Maxwell’s equations solvers to solve the forward problem; our Maxwell’s solvers are new semi-analytical formulations for the computation of the scattered electromagnetic fields from vertically stratified lossy perfect dielectric media. Our Maxwell’s equations solvers are compared with those for a number of cases with known analytical solutions and are shown to be very accurate and efficient. We developed these approaches to overcome some of the limitations and difficulties of the existing approaches(Finite-Difference Time-Domain (FDTD), Finite Element Method (FEM), Boundary Element Method (BEM) …). And then solve the corresponding global minimization problem using simulated annealing algorithm alternativelysolve the inverse problem by training a radial basis neural network.