Abstract: In the application of trust model in WSNs（Wireless Sensor Networks） intrusion detection, the key technologies are studied. A detection model is established, which is mainly about trust index calculation, update and distribution. Update weights are introduced to adjust the ratio of historical and present trust indices, to prevent malicious nodes from accumulating high trust indices. In order to identify malicious nodes and their behaviors, distinguishing mechanism is employed. The simulation results show that the model can resist the middle attacks and other common attacks.

Key words: trust model, update weights, trust calculation, distinguishing mechanism, trust index distribution

摘要： 主要针对信任模型在无线传感器网络入侵检测中应用的关键技术进行了研究，建立以信任值的计算、更新和分发为核心内容的检测模型。为了调节节点的历史行为与目前行为对信任值更新结果的影响，防止恶意节点先积累高信任值然后破坏网络行为的发生，信任计算中引入了更新权值。采用甄别机制以提高对恶意节点及其行为的鉴别。经仿真，结果表明模型可抵御中间人攻击等常见攻击。

关键词: 信任模型, 更新权值, 信任计算, 甄别机制, 信任分发