Secure Node Positioning In Wireless Sensor Networks

Wireless sensor networks have found critical applications in both military and civilian operations where accurate node position information is vital. Location information of sensor is prerequisite not only for location-related applications such as event-monitoring, but for many network functionalities, such as geographical routing. However, the possibility of node compromises and the fact that positioning relies on certain physical features make most of the existing ranging and positioning techniques vulnerable to a variety of attacks.The goal of this dissertation is to protect node location discovery from malicious attacks. On the basis of state-of-the-art researches, many solutions including robust algorithm, robust ranging, malicious beacon detection and location verification are studied through theoretical analysis and simulations. The main contributions of this dissertation are:(1) The security issues of location discovery are investigated, including challenges, possible security attacks and security requirements. Also, the principles, features and limitations of typical secure positioning methods are discussed, along with the future works.(2) A secure node positioning method based on robust algorithm is presented. This method achieves attack-tolerance by using the prediction residual of a least safe reference set to diagnose whether the references are abnormal or not. The least safe reference set is randomly constructed according to the unbiased estimation of variance.(3) A robust node positioning method is developed, which aims at achieving postioning attack-tolerance by removing malicious beacons. We observe that both internal and external attacks usually introduce internal inconsistency (between claimed position and computed position within a beacon) and external inconsistency (between the malicious beacons and the benign ones), and exploit this observation to identify the malicious beacon. We defeat internal attackers and external attackers by applying different strategies, which not only achieve robustness against attacks but reduce the overhead.(4) A secure RSSI-based node positioning scheme is given, which integrates secure ranging, robust algorithm and location verification. In this scheme, the RSSI achieves robustness using integrity coding of beacon signal and occupation of communication channel. Our solution is suitable for the low-cost wireless sensor networks.(5) A trust valuation model based on experience is proposed to evaluate trust between nodes in mobile networks. In this model, the Evidence Theory is introduced to evaluate direct trust and to combine multilateral experience. The observation-frame is proposed to incorporate time sensitivity of experience. A low-cost recommendation technique based on sleep mechanism is provided for neighboring nodes to share experience. Moreover, a fuzzy similarity measure is given to quantitatively evaluate the recommendation trust. Finally, how to apply the proposed techniques to verify the reported position of node is discussed.