Development of Node Cognitive Model for Identification of Malicious Nodes

Abstract

Malicious nodes have been discovered in both wireless and wired networks. The examination of node activity is critical for securing the network's information. The anomalous intrusion detection mechanism examines the node activity and take appropriate action. The cognitives of the node are used to analyse malevolent nodes in this research. The neural network structure is used to analyse node behaviour. It identifies the node class using euclidean distance as a crucial element.. For attack nodes and attack-free nodes, the euclidean distance is calculated. The node behaviour is developed by analysing the node's characteristics. The node dynamics can be efficiently mapped with  multilayer neural network that receives node attributes as input and computes the node class. The backpropagation procedure is performed by the multilayer-feed-forward neural network, which provides results for binary and multi-class output experiments