Inverse scattering of homogeneous dielectric cylinders using genetic programming

Inverse scattering is an important application that can be used to test geometry optimization methods. Indeed, inverse scattering can be conceived as a proxy for design, since an object is sought that scatters waves in a prescribed fashion. The method presented here for homogenous dielectric inverse scattering used genetic programming in the inverse scattering computation of conducting cylinders. To solve dielectric inverse scattering, material parameters must be optimized. This work considers nonmagnetic dielectric materials so materials are represented with a single complex number. Because dielectric inverse scattering is much more demanding than reconstructing conducting cylinders, only convex scatterers are treated here.