Abstract: To resist the influence of the malicious attacks and the malfunctions of fault nodes in wireless sensor networks(WSNs)on data aggregation， this paper proposes an algorithm of trust-based secure data aggregation for WSNs. The algorithm is based on multi-layer non-uniform clustering network topology to achieve secure data aggregation， which can effectively balance the network energy consumption. The trust evaluation model is established based on the communication behavior and data correlation among the nodes. The dynamic trust integration mechanism and update mechanism are introduced to realize the trust evaluation intra-cluster and inter-cluster. Based on the trust value， a trusted aggregation node is chosen in the cluster to complete data fusion of trusted nodes. Simulation results show that the algorithm can achieve accurate and effective trust evaluation， identify internal malicious nodes， and obtain the data aggregation results with high accuracy.

Key words: security, data aggregation, trust model, wireless sensor network, malicious node