Detection of Position Falsification Attack in VANETs using ACO Mr Gouse baig Mohammad , Dr. Prabhakar Kandukuri
VANETs is a major enabling technology for autonomous and connected vehicle. Vehicles communicate wirelessly with other vehicles, sensors, humans, and infrastructure, thereby improving decision-making based on the information received from its surroundings. Nevertheless, the current application to work with high accuracy, information needs to be authenticated, verified and trustworthy. The most important messages in this network are safety message which are periodically broadcasted for various safety and traffic efficiency related applications such as collision avoidance, intersection warning, and traffic jam detection. However, the primary concern is guaranteeing the trustworthiness of the data in the presence of dishonest and misbehaving peers. Misbehaviour detection is still in their infancy and requires a lot of effort to be integrated into the system. An attacker, who is imitating “gost vehicles” on the road, by broadcasting false position information in the safety message, must be detected and revoked permanently from the VANETs. The objective of this work is analyzing safety message and detecting false position information transmitted by the misbehaving nodes. In the paper, we adopted support vector machine (SVM) on VeReMi data set to detect the misbehaviour. We illustrate that the SVM enhanced high-quality detection of modelled attack patterns. This work may feasible and effective way of detecting such misbehaviour in a real-world scenario of VANETs.