First Stage Detection of Node Capture Attacks in WSN based on Cloud Model

摘要

As a typical application of Internet of Things(IoT), the issue of protecting security of WSN will be an important problem effecting the expansion of IoT. Since the nodes of WSN are always arranged in an open environment, they are easy to be captured by adversaries. Traditional detections of capture always judge the attack by comparing the absence time of nodes with a certain threshold. In actual applications of WSN, the uncertainty of radio environment and transmission ability of nodes incurs problem of determination of threshold. According to the quality-quantity uncertainty conversion ability of Cloud Model, it is practical to calculate the absence time of nodes in a certain area, to establish the cloud of wireless environment. The threshold will be determined by cloud of wireless environment. Then, Cloud-based First Stage Detection (CFSD) is proposed to detect the node capture attacks in WSN. Simulation results verify that this approach can improve the accuracy of detection and decrease the false alarm probability.