Restoration of Antenna Radiation Pattern Using Conditional Generative Adversarial Network

A deep learning model based on the conditional generative adversarial network (c-GAN) is proposed to accelerate the measurement of radiation and scattering patterns. Recently, GAN has been proposed to effectively restore an image from its fragments and blank areas. Here, we further leverage the GAN model to fully restore the antenna radiation pattern with sparse solid angles, thus greatly saving the characterization time and cost. We envision that the proposed GAN-assisted measurement method may be beneficial for various applications, including radar, remote sensing, and optimal deployment of 5G/6G networks.