Visible to the public Estimation of safe sensor measurements of autonomous system under attack

TitleEstimation of safe sensor measurements of autonomous system under attack
Publication TypeConference Paper
Year of Publication2017
AuthorsDutta, R. G., Guo, Xiaolong, Zhang, Teng, Kwiat, K., Kamhoua, C., Njilla, L., Jin, Y.
Conference Name2017 54th ACM/EDAC/IEEE Design Automation Conference (DAC)
Keywordsattack vector, autonomous system, car-follower model, challenge response authentication technique, composability, Computer crime, Cyber-physical systems, delay injection, delays, Estimation, Human Behavior, human factors, Mathematical model, message authentication, Metrics, pubcrawl, Radar, radar antennas, recursive least square algorithm, Resiliency, safe sensor measurements, safety concerns, safety-critical software, security concerns, sensor security, Sensors
AbstractThe introduction of automation in cyber-physical systems (CPS) has raised major safety and security concerns. One attack vector is the sensing unit whose measurements can be manipulated by an adversary through attacks such as denial of service and delay injection. To secure an autonomous CPS from such attacks, we use a challenge response authentication (CRA) technique for detection of attack in active sensors data and estimate safe measurements using the recursive least square algorithm. For demonstrating effectiveness of our proposed approach, a car-follower model is considered where the follower vehicle's radar sensor measurements are manipulated in an attempt to cause a collision.
Citation Keydutta_estimation_2017