Securing Connected & Autonomous Vehicles: Challenges Posed by Adversarial Machine Learning and the Way Forward

Abstract

Connected and autonomous vehicles (CAVs) will form the backbone of future next-generation intelligent transportation systems (ITS) providing travel comfort, road safety, along with a number of value-added services. Such a transformation-which will be fuelled by concomitant advances in technologies for machine learning (ML) and wireless communications-will enable a future vehicular ecosystem that is better featured and more efficient. However, there are lurking security problems related to the use of ML in such a critical setting where an incorrect ML decision may not only be a nuisance but can lead to loss of precious lives. In this paper, we present an in-depth overview of the various challenges associated with the application of ML in vehicular networks. In addition, we formulate the ML pipeline of CAVs and present various potential security issues associated with the adoption of ML methods. In particular, we focus on the perspective of adversarial ML attacks on CAVs and outline a solution to defend against adversarial attacks in multiple settings.

Secure and Robust Machine Learning for Healthcare: A Survey

“…In the recent literature, countermeasures against adversarial attacks are categorized into three classes: (1) modifying model; (2) modifying data; and (3) adding an auxiliary model(s) [163]. A taxonomy of such methods is presented in Figure 7 and are discussed next.…”

Towards Artificial-Intelligence-Based Cybersecurity for Robustifying Automated Driving Systems Against Camera Sensor Attacks

“…Deep learning solutions are used in several autonomous vehicle subsystems in order to perform perception, sensor fusion, scene analysis, and path planning. State-of-the-art and human-competitive performance have been achieved by ML on many computer vision tasks related to autonomous vehicles [9].Over the last years it was demonstrated that ML solutions are vulnerable to certain visual attacks [20] that can cause the autonomous vehicles to misbehave in unexpected and potentially dangerous ways, for example on physical modification of the environment and especially traffic signs [17], [13]. It is considered that in these attacks, modifications are physically added to the objects themselves aiming to make the ML system fail but most humans would not consider it suspicious [8].…”

A Survey on Secure Computation Based on Homomorphic Encryption in Vehicular Ad Hoc Networks

“…Because the Cloud server is untrusted, vehicle data may be illegally accessed, forged, tampered with or discarded in the process of transmission and computation. For example, Qayyum et al [127] analyzed extant adversarial machine learning attacks in VANETs. In order to defend against insecure data analysis, there are some efficient solutions, which are described as follows.…”