A QUADRATIC MODEL FOR INVERSE PROFILING - THE ONE-DIMENSIONAL CASE

In this paper a quadratic model for inverse profiling is introduced. I t extends the range of validity of the linear (Born) method, and at th e same time allows to manage the intrinsic non linearity of the proble m. The inversion of the quadratic operator is based on the minimizatio n of a proper defined functional involving both measured data and unkn owns. The adopted geometrical approach, already used in the context of phase retrieval in antennas applications, avoids the occurrence of lo cal minima. Numerical experiments, showing the good performances of th e proposed method, are also presented.