Zeroth-Order Optimization Attacks on Deep Reinforcement Learning-Based Lane Changing Algorithms for Autonomous Vehicles

Dayu Zhang, Nasser Azad, Sebastian Fischmeister, Stefan Marksteiner

2023

Abstract

As Autonomous Vehicles (AVs) become prevalent, their reinforcement learning-based decision-making algorithms, especially those governing highway lane changes, are potentially vulnerable to adversarial attacks. This study investigates the vulnerability of Deep Q-Network (DQN) and Deep Deterministic Policy Gradient (DDPG) reinforcement learning algorithms to black-box attacks. We utilize zeroth-order optimization methods like ZO-SignSGD, allowing effective attacks without gradient information, revealing vulnerabilities in the existing systems. Our results demonstrate that these attacks can significantly degrade the performance of the AV, reducing their rewards by 60 percent and more. We also explore adversarial training as a defensive measure, which enhances the robustness of the DRL algorithms but at the expense of overall performance. Our findings underline the necessity of developing robust and secure reinforcement learning algorithms for AVs, urging further research into comprehensive defense strategies. The work is the first to apply zeroth-order optimization attacks on reinforcement learning in AVs, highlighting the imperative for balancing robustness and accuracy in AV algorithms.

Paper Citation

in Harvard Style

Zhang D., Azad N., Fischmeister S. and Marksteiner S. (2023). Zeroth-Order Optimization Attacks on Deep Reinforcement Learning-Based Lane Changing Algorithms for Autonomous Vehicles. In Proceedings of the 20th International Conference on Informatics in Control, Automation and Robotics - Volume 1: ICINCO; ISBN 978-989-758-670-5, SciTePress, pages 665-673. DOI: 10.5220/0012187700003543

in Bibtex Style

@conference{icinco23,
author={Dayu Zhang and Nasser Azad and Sebastian Fischmeister and Stefan Marksteiner},
title={Zeroth-Order Optimization Attacks on Deep Reinforcement Learning-Based Lane Changing Algorithms for Autonomous Vehicles},
booktitle={Proceedings of the 20th International Conference on Informatics in Control, Automation and Robotics - Volume 1: ICINCO},
year={2023},
pages={665-673},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0012187700003543},
isbn={978-989-758-670-5},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 20th International Conference on Informatics in Control, Automation and Robotics - Volume 1: ICINCO
TI - Zeroth-Order Optimization Attacks on Deep Reinforcement Learning-Based Lane Changing Algorithms for Autonomous Vehicles
SN - 978-989-758-670-5
AU - Zhang D.
AU - Azad N.
AU - Fischmeister S.
AU - Marksteiner S.
PY - 2023
SP - 665
EP - 673
DO - 10.5220/0012187700003543
PB - SciTePress