Inverse MoM Approach to Near-Field Prediction and RFI Estimation in Electronic Devices With Multiple Radiating Elements

Abstract

An inverse method of moments (MoM) approach is developed to estimate the equivalent electric/magnetic current sources to predict radio frequency interference (RFI) in electronic devices. The proposed method is designed to reconstruct the noise sources in electronic devices containing multiple radiating elements by using a finite metal object carrying electric surface currents. The surface current elements are further applied to calculate the electromagnetic fields on arbitrary observation planes near the device under test (DUT). The method is validated through numerical simulation examples, including electrictype CPU noise, magnetic-type non-planar HDMI connector noise, printed circuit board (PCB) with Wi-Fi and PIFA antennas, and heat sink embedded in the DUT. The results confirm less than 0.4 dB difference in the predicted pattern at the antenna operating frequency when a triangular segmentation is applied in the vicinity of the noise source area, and the equivalent surface current is calculated at the edges of the Rao-Wilton-Glisson (RWG) meshes. The proposed method is a promising 'source reconstruction' candidate for predicting the near-field radiation characteristics of electronic devices with multiple peaks and nulls in the near-field radiation pattern. The method helps estimate the level of RFI by summing up the contributions of all the dipoles of the equivalent noise source derived by solving forward and inverse problems. The RFI results based on the inverse MoM approach are validated by comparing them with those obtained from CST simulation, and good agreement is achieved between the two.