Malicious Node Detection on Vehicular Ad-Hoc Network Using Dempster Shafer Theory for Denial of Services Attack

VANET is the network which monitoring the traffic on the road and share the information with neighbors. Movable nodes (vehicle) and fixed node (RSU) are used in VANET. This network is about the safety purpose of driver and sharing confidential information about traffic and accident. On network some vehicles do not share information, send fake requests and try to break network security. Lots of techniques are present here, to overcome by these attacks. Overcome by these issues we give artificial neural network based technique in which we used self-organized map. For trained our network, we use trace file, and these trace file work as input to self-organized map so that we provide supervised learning to our network. In this paper those malicious vehicles have been identified. We used SOM classifier for the detection of misbehavior node. In this classification the group of malicious nodes being created and for the improvement the use of Dempster-Shafer theory for finding attacker node is applied.