Abstract
In this paper a novel method for investigation of inverse scattering in optical complex mediums is proposed. The proposed method is based on radial basis function Neural Networks (RBFNN) and Genetic Algorithms. Medium discrimination is performed by RBFNN and corresponding medium parameters extraction is done using genetic algorithm. In the proposed method for simplicity the Apodized, chirped and simultaneously Apodized and chirped types of mediums are considered as RBFNN library. First RBFNN discriminates medium type and then genetic algorithm extract precisely all necessary and enough discriminated medium parameters. The proposed method tries to open a new insight to inverse scattering in optical devices and systems identification. The simulated results are closely compatible with full numerical simulations to illustrate the ability of the proposed algorithm.