Data-Driven Electromagnetic Scalar Field Estimation of a Patch Antenna Using Deep Convolutional Neural Network
Conference
Khan, MR, Zekios, CL, Bhardwaj, S et al. (2021). Data-Driven Electromagnetic Scalar Field Estimation of a Patch Antenna Using Deep Convolutional Neural Network
. 1497-1498. 10.1109/APS/URSI47566.2021.9704289
Khan, MR, Zekios, CL, Bhardwaj, S et al. (2021). Data-Driven Electromagnetic Scalar Field Estimation of a Patch Antenna Using Deep Convolutional Neural Network
. 1497-1498. 10.1109/APS/URSI47566.2021.9704289
Artificial neural network (ANN) is emerging as an alternative approach for numerical electromagnetic (EM) antenna modeling. In this paper, we focus on predicting the near-field properties of antennas using a data-driven approach, and exploiting the universal function approximator feature of the ANNs. Specifically, we demonstrate the use of a convolutional neural network (CNN) to estimate the surface current on a patch antenna. This field analysis prediction allows a pathway for the prediction of near-field and far-field properties and has the potential to replace full EM modeling. Based on our results, our neural network predicted with only an 11% error on average the desired surface currents, showing promise for the proposed approach.
