SecForCARs logo

Cooperative Adaptive Cruise Control – Part 1: Overview and Attacks

Ulm University, TU Braunschweig

In this demonstrator, we investigate how control systems in automated vehicles can be attacked. We explore this with a relatively simple automated driving function, namely Cooperative Adaptive Cruise Control (CACC). In CACC, vehicles periodically communicate their current driving state (like position, speed, heading, acceleration, or deceleration) to other vehicles with which they form a platoon of vehicles. A CACC control algorithm running in each vehicle uses this information to determine the necessary longitudinal acceleration or deceleration that is required to maintain a stable platoon. Compared to the better-known Adaptive Cruise Control (ACC) that typically relies on RADAR-based distance measurements to the vehicle in front, CACC has the advantages that a better string-stability and shorter driving distances can be achieved.


Attackers that are part of the platoon or can spoof valid messages from platoon members are able to send incorrect information to the other platoon members. It was shown that a CACC control algorithm that is fed with such maliciously manipulated input can lead to instable platoons or even accidents. SecForCARs investigates how misbehavior detection and secure control algorithms can avoid such risks.

Our demonstrator 1 illustrates all these aspects by means of model cars running on a track. For obvious reasons, such a demonstrator cannot be implemented using real-world cars, as one purpose is to provoke accidents. The fact that these model cars can only accelerate or decelerate but otherwise have to follow a fixed track is not a drawback, as CACC is only about longitudinal control anyways.

The videos for this demonstrator are split in multiple parts. Part one explains the setup of the demonstrator and shows the effects of attacks. Future videos will explain how misbehavior detection can help to mitigate attacks and how misbehavior detection related to in-vehicle intrusion detection.