Data-Injection-Proof Predictive Vehicle Platooning: Performance Analysis with Cellular-V2X Sidelink Communications
Siyu Fu,Zhiyuan Jiang,Shunqing Zhang,Shugong Xu,Bin Han,Hans D. Schotten
DOI: https://doi.org/10.1109/jiot.2021.3122125
IF: 10.6
2021-01-01
IEEE Internet of Things Journal
Abstract:The increasing demand for road freight has raised tremendous attention to vehicle platooning, which reduces air resistance and improves fuel economy. To achieve a small and safe spacing between vehicles while ensuring platoon stability, wireless communication assistance is indispensable. However, the imperfection of communication brings degradation of platooning performance (i.e., spacing error) and more possibilities for adversarial attacks. This article proposes a prediction-assisted platooning mechanism from the perspective of performance optimization, wherein each vehicle establishes its local platoon model based on the information received from the communication network, thereby reducing information latency. Then, to secure the system against malicious vehicles, we carry out analysis and design of a detection algorithm for a typical attack type, i.e., the data-injection attack. The detection is based on three indicators: 1) absolute spacing error; 2) spacing prediction error; and 3) acceleration prediction error. The advantages of the novel platooning mechanism and detection algorithm are ultimately demonstrated on a road-traffic simulation platform that considers the imperfection of realistic vehicle perceptions and Cellular-Vehicle-to-Everything (C-V2X) communication.
computer science, information systems,telecommunications,engineering, electrical & electronic