Intrusion detection in ad hoc network using machine learning technique

Ad hoc network is a temporary self-organizing and infrastructure less network. So, it is mostly applied in the military field and disaster relief. Due to wireless communication and self-organizing property ad hoc network is more vulnerable to several intrusions or attacks than the traditional system. Blackhole attack is an important routing disruption attack that malicious node advertises itself as part of a path to the destination. In this paper, we have simulated blackhole attack in ad hoc network environment and collected data of essential features for attack behaviors classification. Then, many machine learning techniques have applied for classification of benign and malicious packet information. It suggests a new approach for select features, essential information collection, and intrusion detection in ad hoc network using machine learning techniques. We have shown comparative results of different machine learning techniques. Our results indicate that this approach can use with different classifiers and can extend it with other intrusions.