Detection and mitigation of sinkhole attacks in wireless sensor networks

• A geostatistical hazard model has been proposed to estimate energy holes in wireless sensor networks.
• We introduce a centralized model to detect sinkhole attacks based on the hazard model.
• We propose a distributed monitoring approach to detect energy holes in the network.
• Analytical models and simulations are provided.

With the advances in technology, there has been an increasing interest in the use of wireless sensor networks (WSNs). WSNs are vulnerable to a wide class of attacks among which sinkhole attack puts severe threats to the security of such networks. This paper proposes two approaches to detect and mitigate such attack in WSNs. It provides a centralized approach to detect suspicious regions in the network using geostatistical hazard model. Furthermore, a distributed monitoring approach has been proposed to explore every neighborhood in the network to detect malicious behaviors. Our simulation experiments validate the correctness and efficiency of the proposed approaches.