Numerical modeling for the localization and the assessment of electromagnetic field sources

In this work, we present a possible approach to electromagnetic source localization using a hybrid blind separation-minimum search inversion algorithm. The total electrical field versus time, emitted by the working antennas located at different and unknown geographical positions, is used, via a suitable blind signal processing technique based on neural networks, to reconstruct each separate contribute. When the emitted electric signals have been separated for each emitting antenna, the unknown locations of the antennas are determined with a minimum-search numerical technique.