On the use of evolutionary algorithms in estimation of permittivity

Evolutionary algorithms used in non-explicit inverse problems may represent a crucial helpful instrument. This work deals with estimation of complex permittivity. Current work widens previous theoretical knowledge and emphasizes on the progress in this field. Backward reconstruction of complex permittivity of single and multi-layered structures in free space in the range of millimetre and sub-millimetre waves remains a challenging and up to date complex problem. The basis of proposed system is represented by a verified direct mathematical model (deduced from Maxwell's equations) computing scattering parameters of defined non-magnetic materials in free space (of known thicknesses and complex permittivities). Turning this explicit model over and thus using transmission coefficients as the input of the system and applying evolutionary algorithm on it (Particle Swarm Optimization has been chosen for this purpose in this case) provides a tool producing useful relatively precise and fast estimations of complex material permittivity.