Abstract
This article reviews a number of algorithms to solve the nonlinear inverse scattering problem, where the discrepancy between the measured data and the predicted data is minimized. These algorithms are all based on a source type of integral representation of the scattered field, where the integral operator acts on a contrast source, being the product of the interior field and the contrast of the scattering object. Special attention is paid to the development of the contrast-source-inversion method, where the contrast sources are updated iteratively, and where, in each iteration, an explicit minimizer for the contrast is obtained. In particular, we discuss the effective implementation of a multiplicative constraint that minimizes the spatial variations of the contrast. We present actual reconstructions from both synthetic and experimental scattered field data.