Deep Learning for Radio Frequency Target Classification

In this lecture, we present modern deep learning (DL) techniques for radio frequency (RF) imagery and signals (i.e., Synthetic Aperture Radar / SAR data, communication signals) classification. First, we will provide a short overview of machine learning (ML) /DL theory and understanding of SAR imagery and RF signals. Then we will demonstrate details algorithmic implementation and performance of DL algorithms on classifying SAR data and RF signals. We will present recent research results, technical challenges, and directions of DL-based object classification for RF sensing. Finally, we will provide adversarial attacks and mitigation techniques involving DL-based RF object recognition.