Detecting packet dropping nodes using machine learning techniques in Mobile ad-hoc network: A survey

Abstract

Mobile ad-hoc networks have to suffer with different types of packet dropping attacks. Therefore, we need strong mechanism to detect these malevolent nodes and to classify normal and abnormal nodes as per the behavior of nodes. Machine learning techniques distinguish outlier nodes quickly and accurately provide classification by observing behavior of those nodes in the network. In this paper, we study various machine learning techniques as artificial neural network, support vector machine, decision tree, Q-learning, Bayesian network for identifying the malicious nodes. These techniques are able to detect black hole, gray hole, flooding attacks and other packet dropping attacks. These types of misbehaving nodes are identified and future behaviors of the nodes are predicted with supervised, un-supervised, reinforcement machine learning techniques.