Malicious Node Detection Using a Dual Threshold in Wireless Sensor Networks

Sensor networks for various event detection applications cannot function effectively if they are vulnerable to attacks. Malicious nodes can generate incorrect readings and misleading reports in such a way that event detection accuracy and false alarm rates are unacceptably low and high, respectively. In this paper, we present a malicious node detection scheme for wireless sensor networks. Unlike others using a single threshold, the proposed scheme employs two thresholds to cope with the strong trade-off between event detection accuracy and false alarm rate, resulting in improved malicious node detection performance. In addition, each sensor node maintains the trust values of its neighboring nodes to reflect their behavior in decision-making. Computer simulation shows that the proposed scheme achieves high malicious node detection accuracy without sacrificing normal sensor nodes and outperforms the scheme using a single threshold.