Parameter inversion of layered media covered by a rough surface

An inversion technique for estimating the structure parameters of layered media covered by a rough surface based on a genetic algorithm is proposed in this work. A theoretical model of scattering from three-dimensional arbitrary layered media with a 3D infinite rough surface based on the small perturbation method (SPM) is employed to calculate the forward scattering problem. The backscatting coefficients of multiple frequencies, angles and polarizations are used to create a nonlinear optimization problem. As the cost function has many local minima, a genetic algorithm is applied to help the inversion procedure converge to the global minimum. The inversion results show the method poses good tolerance to the input data with the additive white Gaussian noise.