Visible to the public An intelligent security system for autonomous cars based on infrared sensors

TitleAn intelligent security system for autonomous cars based on infrared sensors
Publication TypeConference Paper
Year of Publication2017
AuthorsAlheeti, K. M. A., McDonald-Maier, K.
Conference Name2017 23rd International Conference on Automation and Computing (ICAC)
Date Publishedsep
Keywordsautomobiles, autonomous cars, autonomous vehicle system, Autonomous vehicles, awareness messages, composability, control data authentication, external communication systems, feature extraction, Hacking, Human Behavior, human factors, ICMetric technology, ICmetrics, infrared detectors, infrared sensors, Integrated Circuit Metric technology, intelligent security system, intelligent transportation systems, Intrusion detection, Metrics, mobile robots, nonsafety applications, notification messages, pubcrawl, Resiliency, road safety, road vehicles, security of data, security system, self-driving vehicles, sensor security, simulated vehicular ad hoc network, telecommunication security, traffic engineering computing, unique extracted features, vehicular ad hoc networks
AbstractSafety and non-safety applications in the external communication systems of self-driving vehicles require authentication of control data, cooperative awareness messages and notification messages. Traditional security systems can prevent attackers from hacking or breaking important system functionality in autonomous vehicles. This paper presents a novel security system designed to protect vehicular ad hoc networks in self-driving and semi-autonomous vehicles that is based on Integrated Circuit Metric technology (ICMetrics). ICMetrics has the ability to secure communication systems in autonomous vehicles using features of the autonomous vehicle system itself. This security system is based on unique extracted features from vehicles behaviour and its sensors. Specifically, features have been extracted from bias values of infrared sensors which are used alongside semantically extracted information from a trace file of a simulated vehicular ad hoc network. The practical experimental implementation and evaluation of this system demonstrates the efficiency in identifying of abnormal/malicious behaviour typical for an attack.
DOI10.23919/IConAC.2017.8082084
Citation Keyalheeti_intelligent_2017