Intrusion Detection Scheme for Autonomous Driving Vehicles
With the recent breakthroughs, autonomous driving vehicles (ADVs) are promising to bring transformative changes to our transportation systems. However, recent hacks have demonstrated numerous vulnerabilities in these emerging systems from software to cont
- PDF / 467,761 Bytes
- 14 Pages / 439.37 x 666.142 pts Page_size
- 9 Downloads / 288 Views
stract. With the recent breakthroughs, autonomous driving vehicles (ADVs) are promising to bring transformative changes to our transportation systems. However, recent hacks have demonstrated numerous vulnerabilities in these emerging systems from software to control. Safety is becoming one of the major barriers for the wider adoption of ADVs. ADVs connect to vehicular ad-hoc networks (VANETs) to communicate with each other. However, malicious nodes can falsify information and threaten the safety of passengers and other vehicles with catastrophic consequences. In this work, we present a novel reputation-based intrusion detection scheme to detect malicious ADVs through dynamic credit and reputation evaluation. To further encourage user’s participation, an incentive mechanism is also built for ADVs in the intrusion detection system. We demonstrate the feasibility and effectiveness of our proposed system through extensive simulation, compared with current representative approaches. Simulation results show that our proposed scheme can acquire better intrusion detection results, reduced false positive ratio, and improved user participation. Keywords: Autonomous driving vehicles · Credit threshold · Intrusion detection · Incentive model
1
· Dynamic
Introduction
With the rapid development of the automobile industry, many believe that autonomous driving vehicles (ADVs) will bring transformative changes to our society. As a networked cyber-physical system, ADVs will greatly improve the current traffic environment and bring convenience to people’s travel. A number of leading carmakers, including Toyota and Volkswagen, have announced their plain to commercialize self-driving cars to the general public in the next five years. As reported in [1], 25% of the vehicles on the road will be ADVs by 2035. ADVs communicate with each other through vehicular ad-hoc network (VANET), which can be broadly considered as a mobile ad-hoc network on the road tailored for automobiles [2]. For example, ADVs can share information about current road conditions to help others plan their routes. Apart from vehicle-to-vehicle communication, ADVs can also communicate with Roadside c Springer Nature Singapore Pte Ltd. 2020 S. Yu et al. (Eds.): SPDE 2020, CCIS 1268, pp. 278–291, 2020. https://doi.org/10.1007/978-981-15-9129-7_20
Intrusion Detection Scheme for Autonomous Driving Vehicles
279
Units (RSUs) to obtain desired information. All transmitted data which ADVs rely heavily on for safe driving should be protected by the security mechanisms [3]. Existing security approached mainly focus on cryptographic mechanisms. Although cryptographic mechanisms can protect the confidentiality and integrity of data, they cannot cope with the insider attackers well. When there are malicious nodes inside the VANET, they can manipulate data and connection to sabotage the network. For example, some ADVs will release a lot of false news (such as traffic accidents, road congestion, etc.), and these false news may cause chaotic traffic and accidents [4]. Besides, there are
Data Loading...
