Vulnerability Analysis of Vehicular Coordinated Maneuvers

Intelligent Transport Systems (ITS) latest standardization efforts focus on a Maneuver Coordination Service (MCS), for automated vehicles to cooperatively perform maneuvers. The goal is to avoid degrading to lower levels of automation, i.e., human input for maneuvering, e.g., when an obstacle ahead needs to be avoided. MCS-equipped vehicles communicate with nearby vehicles that are possibly affected by the impending maneuver, to establish that a maneuver can safely take place. An MCS-equipped vehicle that misbehaves can be catastrophic: transmitting falsified MCS messages or preventing their reception can mislead victim vehicles into aborting a maneuver, being delayed and, worse even, collide. In this work, we investigate the robustness of existing Maneuver Coordination Protocols (MCPs) and analyze the effect of falsification and jamming attacks. Our analysis shows an increased probability for neck injuries, i.e., whiplash, and potentially more severe injuries. As a first step towards thwarting attacks targeting MCPs, we extend MCPs to take into account on-board vehicle sensors, along with MCP messaging, before committing to a maneuver. Our results demonstrate the MCP vulnerability, the improvement thanks to the sensors, and the need to further improve MCP security. We conclude with a road-map towards a resilient MCS.