Trust-Based Data Aggregation for WSNs in the Presence of Faults and Collusion Attacks

Trust and reputation systems are widely employed in distributed systems, such as sensor networks, to assess the trustworthiness of sources and provide a robust data aggregation. However, some sophisticated attacks, such as node compromise attacks, still can distort computed trust scores, lead to low quality or deceptive services, and undermine users’ confidence. Although fault detection and tolerance problems have been widely studied in trust-based data aggregation systems, taking into account collusion attacks is still a challenging problem.In this dissertation we design and develop data aggregation schemes which leverage a novel trust computation method, robust in the presence of faults and malicious attacks. To this end, we focus on iterative filtering algorithms due to the fact that these algorithms simultaneously aggregate data from multiple sources and provide trust assessment of these sources. We first demonstrate that several iterative algorithms while significantly more robust against node compromise attacks than the simple averaging methods, are nevertheless susceptible to our novel collusion attack. To address this security issue, we propose an improvement for these algorithms which makes them not only collusion robust, but also more accurate and faster converging. We also propose a novel, local, collaborative trust framework, called CrPr, based on an introduced concept of credibility propagation. Furthermore, we augment the proposed trust framework with a novel node compromise detection and revocation method which leverages the error behaviour of the nodes to classify them into benign and compromised nodes. Moreover, given that the original CrPr algorithm may not be efficient for streaming data, we propose an extension to the algorithm, called Str-CrPr which can compute the results in real time.Finally, we propose a novel network security risk assessment which leverages two concepts: (1) an interdependency relationship among the risk scores of network fl