Radiation Source Reconstruction based on Artificial Neural Network for Radio Frequency Interference (RFI) Analysis in Complex System

Abstract

In a modern electronic product, a radio frequency (RF) source usually radiates in a very complex system, where some other components may be mounted near the source causing multi-reflection and/or diffraction effects. In conventional source reconstruction, based on Green’s function in free space, usually omits the effects caused by the nearby components. It may generate an inaccurate result in some practice cases. A new equivalent radiation source reconstruction method based on artificial neural network (ANN) is proposed in this paper. By virtue of the powerful mapping ability of the ANN, such multi-reflection effect can be considered in the training process. The training data is obtained by planar near-field scanning. After obtaining such equivalent source model, we can use it to predict near-fields outside the scanning plane. A numerical example shows that the proposed ANN equivalent source can be better in predicting the shadowing effect of the components around the unknown RF source. This study provides a novel source reconstruction solution to analyze radio frequency interference problems.