Frequency-based Automated Modulation Classification in the Presence of Adversaries

TL;DR

This paper introduces a frequency-domain feature-based deep learning architecture for automatic modulation classification that is robust against transferable adversarial attacks, significantly improving accuracy and security.

Contribution

The work proposes a novel frequency feature-based deep learning model that resists transferable adversarial interference in AMC, outperforming time-domain models.

Findings

Over 30% accuracy improvement on RNNs under attack

Over 50% accuracy improvement on CNNs under attack

Achieves over 99% accuracy without adversarial interference

Abstract

Automatic modulation classification (AMC) aims to improve the efficiency of crowded radio spectrums by automatically predicting the modulation constellation of wireless RF signals. Recent work has demonstrated the ability of deep learning to achieve robust AMC performance using raw in-phase and quadrature (IQ) time samples. Yet, deep learning models are highly susceptible to adversarial interference, which cause intelligent prediction models to misclassify received samples with high confidence. Furthermore, adversarial interference is often transferable, allowing an adversary to attack multiple deep learning models with a single perturbation crafted for a particular classification network. In this work, we present a novel receiver architecture consisting of deep learning models capable of withstanding transferable adversarial interference. Specifically, we show that adversarial attacks crafted to fool models trained on time-domain features are not easily transferable to models trained using frequency-domain features. In this capacity, we demonstrate classification performance improvements greater than 30% on recurrent neural networks (RNNs) and greater than 50% on convolutional neural networks (CNNs). We further demonstrate our frequency feature-based classification models to achieve accuracies greater than 99% in the absence of attacks.