Inverse scattering with far-field intensity data: random surfaces that belong to a well-defined statistical class

Demetrio Macías (1) , Gustavo Olague (2) , Eugenio Méndez (2)

1 LNIO - Laboratoire de Nanotechnologie et d'Instrumentation Optique
2 CICESE - Centro de Investigacion Cientifica y de Education Superior de Ensenada [Mexico]

We consider the inverse scattering problem of retrieving the surface profile function from far-field angle-resolved intensity data. The problem is approached as a nonlinear constrained optimization problem. The surface, assumed one-dimensional and perfectly conducting, is also assumed to be a realization of a Gaussian random process with a Gaussian correlation function with known standard deviation of heights (δ) and correlation length (a). Starting from rigorously calculated far-field angle-resolved scattered data, we search for the optimum profile using evolutionary strategies. Examples that illustrate the proposed scheme are presented. Aspects of the convergence and lack of uniqueness of the solution are discussed.

Domaines: Analyse de données, Statistiques et Probabilités [physics.data-an]

Demetrio Macías, Gustavo Olague, Eugenio Méndez. Inverse scattering with far-field intensity data: random surfaces that belong to a well-defined statistical class. Waves in Random and Complex Media, 2006, 16 (4), pp.545-560. ⟨10.1080/17455030600830385⟩. ⟨hal-02372999⟩