一种基于多维度信任的WSN安全数据融合方法

A Dimensional Trust Based Security Data Aggregation Method in Wireless Sensor Networks

针对无线传感器内部恶意节点可能产生的攻击行为进行分析,提出了基于无核心节点的信任评估机制,邻居节点间互相进行行为监控,利用一种基于多维度信任的计算模型得到了对方的信任值,再根据由D-S证据理论的合成法则,对直接信任与间接信任进行合成.实验结果表明,提出的方法能够有效对恶意节点的重放攻击、数据包数量异常、伪造数据等攻击行为进行限制,降低其节点信任值,提高无线传感网络的安全性和可靠性.

Aiming at internal malicious attack behaviors generated from the wireless sensor node, this paper pres- ents trust evaluation mechanism based on the nodes~ reputation, the neighbor node r：nutual behavior could be moni- tored and used an dimensional trust computational model to calculate each other＇s trust value. According to the combi- nation rule of the D-S evidence theory, the direct and indirect trust could be merged. The experiment shows that this method can effectively restrict the behavior of packets replay attack, abnormal quantity of packets and forgery attack derived from the malicious nodes, and reduce the node trust value and then improve the security and reliability of the wireless sensor network.