Clock Skew Based Node Identification in Wireless Sensor Networks

Node identification is one of the most important issues to wireless sensor network security. Current approaches use cryptographic authentication and certification tools to ensure the node identification, while this paper introduces an intuitive method to identify a node by measuring its clock skew. This method is based on our observation that every sensor node has a unique clock skew value that is different from any other node. We adopt Flooding Time Synchronization Protocol (FTSP) as the measuring tools, and the experimental data show that almost all measured clock skews of one sensor node vary inside a tiny bound. For any two nodes that their clock skews are very close to each other, a classifying function is proposed to check the line continuity of contiguous measured clock skews. The proposed method has successfully identified every node in our experiments, and its applications like Sybil attack detection is also discussed.