Multi-BSM: An Anomaly Detection and Position Falsification Attack Mitigation Approach in Connected Vehicles

Abstract

With the dawn of the emerging technologies in the field of vehicular environment, connected vehicles are advancing at a rapid speed. The advancement of such technologies helps people daily, whether it is to reach from one place to another, avoid traffic, or prevent any hazardous incident from occurring. Safety is one of the main concerns regarding the vehicular environment when it comes to developing applications for connected vehicles. Connected vehicles depend on messages known as basic safety messages (BSMs) that are repeatedly broadcast in their communication range in order to obtain information regarding their surroundings. Different kinds of attacks can be initiated by a vehicle in the network with malicious intent by inserting false information in these messages, e.g., speed, direction, and position. This paper focuses on the position falsification attacks that can be carried out in the vehicular environment and be avoided using the multi-BSM approach. Multi-BSM uses consecutive multiple BSMs with different parameters to detect and warn other vehicles about position falsification attacks. Multi-BSM is compared to other anomaly detection algorithms and evaluated with rigorous simulations. Multi-BSM shows a high level of anomaly detection, even in high vehicle density, with up to 97% accuracy rate compared to the respective algorithms.