A Mechanism to Classify The Black Hole Behavior of Vehicles in Vanets using One Class SVM
Due to the nature such as the absence of fixed infrastructure, VANets are known to be very vulnerable to various
internal and external attacks. Particularly, one of very common attack in vehicular ad hoc network is black hole attack. In
this paper, an anomaly based mechanism is proposed to classify and detect the behavior of vehicles in Vanet. SUMO and ns2
to are used to simulate vehicular network and generate trace files. The system uses the features obtained from the trace file to
detect the normal and abnormal behavior of vehicles using one class Support Vector Machine.
Index Terms- Anomaly detection, Black hole attack, one-class SVM, Security, Vanet.