Improving the resilience of a CACC controller against jamming attacks

Abstract

This research explores how the resilience of a particular Cooperative Adaptive Cruise Controller (CACC) algorithm "P1" can be improved, specifically examining the impact of wireless jamming attacks on Vehicle-to-Vehicle (V2V) communication and implementing rudimentary countermeasures. Previous research has shown that there are vulnerabilities in this P1 algorithm and based on this research we propose and implement various methods to enhance its resilience and finally compare these results to the original algorithm. We utilize the ComFASE simulation environment, which integrates existing simulation frameworks such as Plexe, Veins, SUMO and OmNET++ to perform vehicular and network dynamics simulations. ComFASE also adds on a layer of fault and attack injection tools which we use in conjunction with the frameworks to inject barrage jamming attacks, design countermeasures and finally carry out another round of attack injections to gauge their efficacy. The P1 controller algorithm described further in our thesis was found to cause collisions when under attack in 1475 out of the 3575 experiments we ran. In order to improve its resilience, we propose fallback models as countermeasures. Our best proposed fallback called Model 4c managed to reduce the number of collisions to 637 out of the same 3575 experiments. With this simple yet effective fallback mechanism, we have managed to get a collision reduction of 56.81% when covering a broad range of noise values. If we only consider noise levels greater than 0.66 mW, we have a close to 100% reduction in number of collisions. These results show that the resilience of the P1 controller against barrage jamming attacks can be significantly improved even with just relatively simple modifications to existing algorithms.