Investigation of near field data sampling approaches for far field radiation prediction of PCBs by genetic algorithm

Near field data sampling techniques, including where to place observation points and how to pre-process the data, are investigated in respect to a Near Field-Far Field transformation. For a simple PCB, the optimum equivalent set of basic radiators is found through a Genetic Algorithm. The magnetic near field amplitudes for different observation surfaces of the PCB are obtained from Method of Moment simulations, and the far field patterns for different frequencies are predicted.