Assessing the Impact of Attacks on Connected and Autonomous Vehicles in Vehicular Ad Hoc Networks

Abstract

The transportation sector is evolving rapidly towards more sustainable and safer solutions with the idea of Connected and Autonomous Vehicles (CAVs) based upon Vehicular Ad-hoc Network technology. The biggest challenge for CAVs is the security threats due to their open nature and internet connections, opening a wide range of vulnerabilities. In this research, the impact of four cyber security attacks (Distributed Denial of Service (DDoS), Man-in-the-Middle (MITM), Blackhole and Grayhole) is quantified in terms of network and transportation performance metrics. The map is setup based on a busy urban area in a UK city, and a combination of OMNeT++, Sumo and Veins software tools are used for modelling and simulating the attacks on the network. The simulation is performed with and without the attacks for an accident scenario. MITM is found to have maximum impact severity on the transportation operational efficiency and safety of the CAV network. The dynamic rerouting algorithm of the network is identified as the most vulnerable attack vector, experiencing maximum impact from all the attacks. A maximum packet loss of 82% is achieved by a DDoS attack. These insights showcased the importance of analysing the impacts of security attacks on the transportation efficiency of the CAV network, which is vital for building reliable and safer next-generation mobility systems.