Single Hopdetection Of Node Clone Attacks In Mobile Wireless Sensor Networks

Wireless sensor networks (WSNs) are networks consisting of sensor nodes deployed in an area for data gathering and processing of that deployment area, by means of a variety of computing, communication, sensor and distributed processing technologies. Sensor nodes in WSNs are able to get organized autonomously, gather data and monitor the environment cooperatively. WSNs now see an increasing range of applications in areas like environmental monitoring, wildlife tracking, forest fire forecasting, military surveillance, etc.WSNs are often deployed in hostile and unattended environments. Since low-cost sensor nodes are often not tamper-resistant, they are vulnerable to physical capture and compromise. In a node replication attack or a node clone attack, an adversary physically captures a sensor node, extracts cryptographic secrets from the node, and distributes a large number of replicas of captured node throughout the network. Since replicas carry the extracted cryptographic secrets of the captured node, they will be recognized as legitimate members of the network by other sensor nodes. Therefore, the replicas could be used to launch a variety of insider attacks.Existing node clone attack detection schemes are mainly focused on stationary sensor networks, relying on the fact that locations of sensor nodes do not change after deployment. They won't work in mobile WSNs, where sensor nodes are expected to move around in the deployment area. What's more, most existing detection schemes make unrealistic assumptions on the ability of replicas; they fail to hold if replicas collude to cheat the detection protocol in order to subvert the detection protocol. Also the replicas can turn to attack the routing protocol if reliable communication between distant nodes is required in the clone detection protocol. In this paper, we propose a novel node clone attack detection protocol, namely the Single Hop Detection (SHD) protocol, which is fully distributed, highly robust against node colluding, specifically designed for mobile WSNs. Also we introduce the concept of node revocation process, namely the revocation of detected replicas, which lacks attention in previous works. What is problematic in the revocation process lies in the fact that the revocation process itself might be exploited by the adversary to evict"clean"nodes from the network. In this paper, a simple yet highly efficient solution to tackle this problem will be proposed.Also we give analytical performance evaluation of SHD protocol. With our protocol, the detection of node clone attack is effectively deduced to be an extended version of mobility-assisted routing problem, effectively bridging two different research areas. We show analytically under popular mobility models such as Random Direction, Rand Waypoint and Random Direction mobility models that our protocol provides effective, efficient and robust node clone attack detection capability.