Trust management has an important significance on the reliable operation and safety guarantee in Wireless Sensor Networks(WSNs). Previous works were implemented by exchanging key among neighboring sensors. Considering that in many cases, the sensor nodesȁ9; behavior is both spatio-temporal correlated, and this paper proposes a reputation model called BRMSN (Behavior Reputation Method for Sensor Networks) which focuses on local testing. The model defines the similarity and the similarity matrix by using normal differences among status estimate vectors (or measured values). Using the homogeneity measures that formed by the spatial information between nodes as indirect reputation measurement. Using the reliability measures that formed by the temporal information between nodes as direct reputation measurement. The simulation results show that the model can reduce the number of data exchange among sensors, also our model has the powerful ability to prevent from malicious attacks or faulty nodes.
展开▼
