Use of polarization in electromagnetic inverse scattering


  • Wolfgang-M. Boerner


The complete description of electromagnetic scattering processes implies polarization and since an electromagnetic scatterer acts like a polarization transformer we require measurements for the complete description of the target scattering matrices so that the descriptive parameters of a scatterer can be uniquely recovered from the measured field data. For the purpose of introducing the concept of polarization utilization in practice, the radar case is chosen and generalizations to other electromagnetic inverse problems are presented in the conclusions. For example, in radar target discrimination, identification and imaging use of measurement data available over the entire spatial frequency domain of the radar cross section must be made leading to various approximate frequency domain related approaches. Mainly for historical reasons of having had amplitude data available only, in most cases the approximations had been simplified to purely scalar nature; i.e., polarization-dependent properties were discarded, and the resulting theories are no longer valid or unique. It is the main objective of this analysis to show that because of the vector nature of electromagnetic inverse scattering, theories, if applicable in practice, require incorporation of complete polarization information into their formulation. By applying this approach to existing theories, it is shown that remarkable improvements in fidelity and quality of the reconstructed images are obtained and that indeed there is ample justification for continuing efforts in developing methods and theories of inverse scattering applicable to all those fields of physical sciences where information on the characteristic parameters of a scattering process is to be drawn from remote measurements, be it the electromagnetic vector case or the even more complicated seismic case of s and p wave interactions in elastic media.