Adversarial Robustness of Distilled and Pruned Deep Learning-based Wireless Classifiers

arXiv:2404.15344v1 Announce Type: cross
Abstract: Data-driven deep learning (DL) techniques developed for automatic modulation classification (AMC) of wireless signals are vulnerable to adversarial attacks. This poses a severe security threat to the DL-based wireless systems, specifically for edge applications of AMC. In this work, we address the joint problem of developing optimized DL models that are also robust against adversarial attacks. This enables efficient and reliable deployment of DL-based AMC on edge devices. We first propose two optimized models using …

abstract adversarial adversarial attacks applications arxiv attacks classification classifiers cs.cr cs.it cs.lg data data-driven deep learning edge eess.sp math.it robustness security stat.ml systems threat type vulnerable wireless work