/Security-and-Robustness-of-Deep-Learning-in-Wireless-Communication-Systems

A research oriented repository on the Security and Robustness of Deep Learning for Wireless Communication Systems

Primary LanguageJupyter Notebook

Security and Robustness of Deep Learning in Wireless Communication Systems

This repository is a research oriented-repository dedicated to the Security and robustness of deep learning for wireless communication systems.

The main idea is to release the code of our papers in a very friendly manner such that it cultivates and accelerates further research on this topic. The codes are released in a series of Jupyter notebooks, which are developed with a how to do it in a step-by-step approach. We also provide a research library dedicated to the latest works on this topic.

If you have any comments, please contact me via m(dot)sadeghee(at)gmail(dot)com.

I hope you enjoy it and good luck with your research.

Citing this Repo

If you use any part of this repo, please consider citing our following works:

  • M. Sadeghi and E. G. Larsson, "Physical Adversarial Attacks Against End-to-End Autoencoder Communication Systems," in IEEE Communications Letters, 2019, to appear.
  • M. Sadeghi and E. G. Larsson, “Adversarial attacks on deep-learning based radio signal classification,” IEEE Wireless Communications Letters, vol. 8, no. 1, pp. 213–216, Feb. 2019.

1. Our Works and Our Codes

1.1. Physical Adversarial Attacks Against End-to-End Autoencoder Communication Systems

  • The paper is available at [Paper]
  • The main code is available at [Code]
  • We also have done some extra simulations considering some variations of the set-up we used in our paper. These variations are:
    • Considering other information rates [Code]
    • Using ReLU instead of eLU and also deeper networks [Code]
    • Averaging over different models rather than using one single model [Code]
  • About the work: This work is a follow up of our previous work which presents two new directions. First, it presents the concept of physical attacks in wireless communication systems. Second, it introduces adversarial attacks against end-t-end autoencoder systems.

1.2. Adversarial Attacks on Deep-Learning Based Radio Signal Classification

  • The paper is available at [Paper]

  • The code is available at [code]

  • About the work: This work establishes the concept of adversarial attacks in wireless systems. It shows that DL based wireless applications can be very vulnerable to this new kind of threat. We establish the main concepts, explain how to formulate and also adapt this new concept in the context of wireless systems, and present some novel algorithms for crafting white-box and also black-box attacks.



2. Papers on Security and Robustness of Deep Learning in Wireless Communication Systems

Below is a list of works studying security and robustness in the context of wireless communication. If you want to add your paper here, please send me an email.

  • M. Sadeghi and E. G. Larsson, “Adversarial attacks on deep-learning based radio signal classification,” IEEE Wireless Communications Letters, vol. 8, no. 1, pp. 213–216, Feb. 2019.
  • M. Sadeghi and Erik G. Larsson , “Physical Adversarial Attacks Against End-to-End Autoencoder Communication Systems,” IEEE Communications Letters, 2019.
  • Y. Shi, Y. E. Sagduyu, T. Erpek, K. Davaslioglu, Z. Lu and J. H. Li, “Adversarial deep learning for cognitive radio security: Jamming attack and defense strategies,” in Proc. IEEE International Conference on Communications Workshops (ICC Workshops), 2018.
  • T. Erpek, Y. E. Sagduyu and Y. Shi, “Deep learning for launching and mitigating wireless jamming attacks,” in IEEE Transactions on Cognitive Communications and Networking., 2018.
  • K. K. Nguyen, D. T. Hoang, D. Niyato, P. Wang, D. Nguyen and E. Dutkiewicz, “Cyberattack detection in mobile cloud computing: A deep learning approach,” in Proc. IEEE Wireless Communications and Networking Conference (WCNC), 2018.
  • A. Diro and N. Chilamkurti, “Leveraging LSTM networks for attack detection in fog-to-things communications,” in IEEE Communications Magazine, vol. 56, no. 9, pp. 124-130, Sept. 2018.
  • I. Shakeel, “Machine learning based featureless signalling,” in Proc. IEEE Military Communications Conference (MILCOM), October 2018.
  • F. B. Mismar and B. L. Evans, “Deep Q-Learning for self-organizing networks fault management and radio performance improvement,” to appear in Proc. Asilomar Conference on Signals, Systems, and Computers, October 2018.
  • Y. Shi, T. Erpek, Y. E. Sagduyu and J. H. Li, “Spectrum data poisoning with adversarial deep learning,” in Proc. IEEE Military Communications Conference (MILCOM), 2018.