Classification approach to avoid link failures in wireless sensor networks in mobile virtual communities and teleworking

Security issues are the primary issues in wireless sensor networks (WSN) due to its large network coverage and number of nodes. The attackers in WSN attack the particular node which has a low energy level and converts this node into malicious node. The formation of malicious node is the primary reason for link failures between nodes. This paper proposes an efficient methodology to detect the malicious nodes in WSN using feed forward back propagation neural network classifier. This classifier differentiates the malicious node from trusty node based on the extracted features of the test node. The performance of the proposed malicious node detection system is analysed in terms of detection rate, packet delivery ratio (PDR) and latency. The experimental results are compared with state-of-art methods.