A Method to Extract Physical Dipoles for Radiating Source Characterization and Near Field Coupling Estimation

TLDR: In this article, a hybrid method for physical dipole extraction based on genetic algorithm and linear least square method is proposed, which offers an automatic flow to extract the equivalent dipoles without prior decision of the type, position, orientation and number of dipoles.

Abstract: Infinitesimal electric and magnetic dipoles are widely used as an equivalent radiating source model in far field radiated emissions and near field coupling scenarios. In this paper, a hybrid method for physical dipole extraction based on genetic algorithm and linear least square method is proposed. It offers an automatic flow to extract the equivalent dipoles without prior decision of the type, position, orientation and number of dipoles. Compared with conventional linear least square method, this algorithm can extract physical dipoles which are close to original radiating source and minimize the number of dipoles. Compared with conventional genetic algorithm based method, this method reduces the optimization time and is more robust. This method is validated by both simulation and measurement data, and its advantages are proved. It is applied to modeling of the radiation from a clock buffer chip. The extracted equivalent dipoles are used to estimate the near field coupling from the clock buffer chip to a victim inverted F antenna (IFA) in a practical printed circuit board (PCB) by full wave simulation. The estimation matches well with measurement.