VND-NN: Neural Network Based Vampire Node Detection in Wireless Sensor Networks

Abstract

Wireless sensor networks are widely used for environment monitoring. The sensor nodes are deployed in the areas of interest to obtain the physical world data. These nodes are resource constrained in power, memory and computation. Boarder security is one of the major applications of WSNs. The sensor nodes are mounted with the environmental sensors (temperature, pressure etc.,) to monitor the environment and forward the sensed data to the base station via mobile sinks. Adversary can easily capture nodes from the network because they are not tamper proof in nature. Once the attacker succeed in capturing a legitimate sensor node, can introduce a variety of attacks. Vampire attack a potential attack in WSNs, which is a resource draining attack. The compromised nodes will misbehave with the legitimate nodes in the network and tries to drain the resources continuously resulting in network collapse. In this paper a neural network based misbehavior node detection algorithm is proposed. The simulation result shows that the proposed model effectively detects the misbehavior nodes and conserves the network resources.