Establishing a secure routing in MANET using a hybrid intrusion detection system

MANET has lead to the emergence of on the fly networking among the communicating nodes. However, they are vulnerable to a variety of attacks, especially in the routing layer because of their open environment and the communication happens based on the mutual trust among the nodes. An appropriate intrusion detection system is an inevitable requirement. Using anomaly based intrusion detection system results in the false positive identification of the intrusion. Computing trust value for each node in the network leads to higher computational complexity. Combining trust value along with intrusion detection system can reduce the pitfalls of both methods. A comparative study of using only anomaly based intrusion detection system and combination of trust value along with an anomaly based intrusion detection system is performed. The proposed anomaly detection system uses artificial neural networks and Dempster-Shafer theory to confirm the attack occurrence in the network.