Cybersecurity of Autonomous Vehicles: A Systematic Literature Review of Adversarial Attacks and Defense Models

Author(s):

Mansi Girdhar Junho Hong John Moore

Journal

Institute of Electrical and Electronics Engineers (IEEE)

Abstract

Autonomous driving (AD) has developed tremendously in parallel with the ongoing development and improvement of deep learning (DL) technology. However, the uptake of artificial intelligence (AI) in AD as the core enabling technology raises serious cybersecurity issues. An enhanced attack surface has been spurred on by the rising digitization of vehicles and the integration of AI features. The performance of the autonomous vehicle (AV)-based applications is constrained by the DL models' susceptibility to adversarial attacks despite their great potential. Hence, AI-enabled AVs face numerous security threats, which prevent the large-scale adoption of AVs. Therefore, it becomes crucial to evolve existing cybersecurity practices to deal with risks associated with the increased uptake of AI. Furthermore, putting defense models into practice against adversarial attacks has grown in importance as a field of study amongst researchers. Therefore, this study seeks to provide an overview of the most recent adversarial defensive and attack models developed in the domain of AD.

Concluding remarks

In order to help human drivers in specific situations, such as preventing lane drift or assisting the driver in coming to a complete halt in time to avoid a potential AV crash, the thriving automated vehicles currently come equipped with ADAS safety capabilities. The future L4 and L5 vehicles, on the other hand, will be equipped with advanced ADS functionality, which relies solely on sophisticated AI/ML algorithms for sensor processing, sensor fusion, scene creation, and motion planning, hence allowing the vehicle to sense its environment and make informed decisions. This article has therefore outlined the importance of ML and DL technologies in AVs in, for instance, object classification and detection, SS, etc. However, recent investigations and groundwork on adversarial attacks elicit suspicions about AVs' security and potency. It is inferred that the uptake of these advanced AI technologies has introduced a range of vulnerabilities in the AVs featuring them. This article presented a detailed investigation of adversarial attacks on the automated driving ecosystem. In addition, a survey of adversarial defense models for improving resistance against adversarial attacks on ML components in AVs has been discussed. In order to strengthen resilience and secure the ML and training data for AVs, the authors have especially addressed the significance of generic defense techniques. Finally, the authors expect that the academic community will use this study to help discover any loopholes and withholdings in the existing studies on adversarial attacks and defense strategies for ADSs.

Reference details

DOI

10.1109/OJVT.2023.3265363

Resource type

Journal Article

Year of Publication

2023

ISSN Number

2644-1330

Publication Area

Cybersecurity and defense

Date Published

2023

URL

https://ieeexplore.ieee.org/ab…

How to cite this reference:

Girdhar, M., Hong, J., & Moore, J. (2023). Cybersecurity of Autonomous Vehicles: A Systematic Literature Review of Adversarial Attacks and Defense Models. Institute of Electrical and Electronics Engineers (IEEE). https://doi.org/10.1109/OJVT.2023.3265363 (Original work published 2023)