Adversarial attacking and defensing modulation recognition with deep learning in cognitive radio-enabled IoT

Zhang, Z.; Ma, L.; Liu, M.; Chen, Y.; Zhao, N.

Abstract

Modulation recognition using deep learning (DL) can efficiently recognize modulated signals in cognitive radio-enabled Internet of Things (IoT). However, it is vulnerable to the attack of adversarial examples designed by attackers, leading to a decrease in its accuracy. Different adversarial techniques can be used for attacks, but these attacks have limited efficiency. This paper proposes a double loop iterative method. Different from the traditional attack methods, the new method designs an additional external loop iteration for high efficiency. When generating adversarial examples, the initial conditions of each iteration can be updated as the number of iterations changes, so that the adversarial examples can cross the decision boundary of the model as much as possible. In addition, this paper uses knowledge distillation to improve the traditional adversarial training defense, which improves the robustness of the model. Simulation results show that the proposed attack and defense methods have better performance than traditional methods.

Citation

Zhang, Z., Ma, L., Liu, M., Chen, Y., & Zhao, N. (2024). Adversarial attacking and defensing modulation recognition with deep learning in cognitive radio-enabled IoT. IEEE Internet of Things Journal, 11(8), 14949-14962. https://doi.org/10.1109/JIOT.2023.3345937

• Journal Article Type: Article
• Acceptance Date: Dec 18, 2023
• Online Publication Date: Dec 22, 2023
• Publication Date: Apr 15, 2024
• Deposit Date: Dec 20, 2023
• Publicly Available Date: Dec 20, 2023
• Journal: IEEE Internet of Things Journal
• Electronic ISSN: 2327-4662
• Publisher: Institute of Electrical and Electronics Engineers
• Peer Reviewed: Peer Reviewed
• Volume: 11
• Issue: 8
• Pages: 14949-14962
• DOI: https://doi.org/10.1109/JIOT.2023.3345937
• Public URL: https://durham-repository.worktribe.com/output/2048693