Defense Strategies Against Adversarial Jamming Attacks via Deep Reinforcement Learning

Feng Wang; Chen Zhong; M. Cenk Gursoy; Senem Velipasalar

doi:10.1109/CISS48834.2020.1570629719

Defense Strategies Against Adversarial Jamming Attacks via Deep Reinforcement Learning

Feng Wang, Chen Zhong, M. Cenk Gursoy, Senem Velipasalar

Research output: Chapter in Book/Entry/Poem › Conference contribution

21 Scopus citations

Abstract

As the applications of deep reinforcement learning (DRL) in wireless communication grow, sensitivity of DRL based wireless communication strategies against adversarial attacks has started to draw more attention. In order to address such sensitivity and alleviate the resulting security concerns, we in this paper study defense strategies against DRL-based jamming attacker on a DRL-based dynamic multichannel access agent. To defend the jamming attacks, we propose three diversified defense strategies: proportional-integral-derivative (PID) control, the use of an imitation attacker and the development of orthogonal policies. We design these strategies and evaluate their performances.

Original language	English (US)
Title of host publication	2020 54th Annual Conference on Information Sciences and Systems, CISS 2020
Publisher	Institute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)	9781728140841
DOIs	https://doi.org/10.1109/CISS48834.2020.1570629719
State	Published - Mar 2020
Event	54th Annual Conference on Information Sciences and Systems, CISS 2020 - Princeton, United States Duration: Mar 18 2020 → Mar 20 2020

Publication series

Name	2020 54th Annual Conference on Information Sciences and Systems, CISS 2020

Conference

Conference	54th Annual Conference on Information Sciences and Systems, CISS 2020
Country/Territory	United States
City	Princeton
Period	3/18/20 → 3/20/20

Keywords

Defense strategies
adversarial policies
deep reinforcement learning
dynamic channel access

ASJC Scopus subject areas

Computer Networks and Communications
Information Systems
Signal Processing
Information Systems and Management
Safety, Risk, Reliability and Quality
Artificial Intelligence

Access to Document

10.1109/CISS48834.2020.1570629719

Cite this

Wang, F., Zhong, C., Gursoy, M. C., & Velipasalar, S. (2020). Defense Strategies Against Adversarial Jamming Attacks via Deep Reinforcement Learning. In 2020 54th Annual Conference on Information Sciences and Systems, CISS 2020 Article 9086263 (2020 54th Annual Conference on Information Sciences and Systems, CISS 2020). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/CISS48834.2020.1570629719

Defense Strategies Against Adversarial Jamming Attacks via Deep Reinforcement Learning. / Wang, Feng; Zhong, Chen; Gursoy, M. Cenk et al.
2020 54th Annual Conference on Information Sciences and Systems, CISS 2020. Institute of Electrical and Electronics Engineers Inc., 2020. 9086263 (2020 54th Annual Conference on Information Sciences and Systems, CISS 2020).

Research output: Chapter in Book/Entry/Poem › Conference contribution

Wang, F, Zhong, C, Gursoy, MC & Velipasalar, S 2020, Defense Strategies Against Adversarial Jamming Attacks via Deep Reinforcement Learning. in 2020 54th Annual Conference on Information Sciences and Systems, CISS 2020., 9086263, 2020 54th Annual Conference on Information Sciences and Systems, CISS 2020, Institute of Electrical and Electronics Engineers Inc., 54th Annual Conference on Information Sciences and Systems, CISS 2020, Princeton, United States, 3/18/20. https://doi.org/10.1109/CISS48834.2020.1570629719

Wang F, Zhong C, Gursoy MC , Velipasalar S. Defense Strategies Against Adversarial Jamming Attacks via Deep Reinforcement Learning. In 2020 54th Annual Conference on Information Sciences and Systems, CISS 2020. Institute of Electrical and Electronics Engineers Inc. 2020. 9086263. (2020 54th Annual Conference on Information Sciences and Systems, CISS 2020). doi: 10.1109/CISS48834.2020.1570629719

@inproceedings{369eb42fb0b84c99982bcc1000ef9a37,

title = "Defense Strategies Against Adversarial Jamming Attacks via Deep Reinforcement Learning",

abstract = "As the applications of deep reinforcement learning (DRL) in wireless communication grow, sensitivity of DRL based wireless communication strategies against adversarial attacks has started to draw more attention. In order to address such sensitivity and alleviate the resulting security concerns, we in this paper study defense strategies against DRL-based jamming attacker on a DRL-based dynamic multichannel access agent. To defend the jamming attacks, we propose three diversified defense strategies: proportional-integral-derivative (PID) control, the use of an imitation attacker and the development of orthogonal policies. We design these strategies and evaluate their performances.",

keywords = "Defense strategies, adversarial policies, deep reinforcement learning, dynamic channel access",

author = "Feng Wang and Chen Zhong and Gursoy, {M. Cenk} and Senem Velipasalar",

note = "Publisher Copyright: {\textcopyright} 2020 IEEE.; 54th Annual Conference on Information Sciences and Systems, CISS 2020 ; Conference date: 18-03-2020 Through 20-03-2020",

year = "2020",

month = mar,

doi = "10.1109/CISS48834.2020.1570629719",

language = "English (US)",

series = "2020 54th Annual Conference on Information Sciences and Systems, CISS 2020",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

booktitle = "2020 54th Annual Conference on Information Sciences and Systems, CISS 2020",

}

TY - GEN

T1 - Defense Strategies Against Adversarial Jamming Attacks via Deep Reinforcement Learning

AU - Wang, Feng

AU - Zhong, Chen

AU - Gursoy, M. Cenk

AU - Velipasalar, Senem

PY - 2020/3

Y1 - 2020/3

N2 - As the applications of deep reinforcement learning (DRL) in wireless communication grow, sensitivity of DRL based wireless communication strategies against adversarial attacks has started to draw more attention. In order to address such sensitivity and alleviate the resulting security concerns, we in this paper study defense strategies against DRL-based jamming attacker on a DRL-based dynamic multichannel access agent. To defend the jamming attacks, we propose three diversified defense strategies: proportional-integral-derivative (PID) control, the use of an imitation attacker and the development of orthogonal policies. We design these strategies and evaluate their performances.

AB - As the applications of deep reinforcement learning (DRL) in wireless communication grow, sensitivity of DRL based wireless communication strategies against adversarial attacks has started to draw more attention. In order to address such sensitivity and alleviate the resulting security concerns, we in this paper study defense strategies against DRL-based jamming attacker on a DRL-based dynamic multichannel access agent. To defend the jamming attacks, we propose three diversified defense strategies: proportional-integral-derivative (PID) control, the use of an imitation attacker and the development of orthogonal policies. We design these strategies and evaluate their performances.

KW - Defense strategies

KW - adversarial policies

KW - deep reinforcement learning

KW - dynamic channel access

UR - http://www.scopus.com/inward/record.url?scp=85085247478&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85085247478&partnerID=8YFLogxK

U2 - 10.1109/CISS48834.2020.1570629719

DO - 10.1109/CISS48834.2020.1570629719

M3 - Conference contribution

AN - SCOPUS:85085247478

T3 - 2020 54th Annual Conference on Information Sciences and Systems, CISS 2020

BT - 2020 54th Annual Conference on Information Sciences and Systems, CISS 2020

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 54th Annual Conference on Information Sciences and Systems, CISS 2020

Y2 - 18 March 2020 through 20 March 2020

ER -

Defense Strategies Against Adversarial Jamming Attacks via Deep Reinforcement Learning

Abstract

Publication series

Conference

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this