摘要
引入了主观信任评价及其期望概率的概念,通过统计方法评估船联网中各个节点的可信程度,结合KL距离计算节点的信誉度,从而建立了基于节点信誉度评价的数据融合信任模型。为避免影响融合的结果,模型在数据融合过程中保留信誉度较好的节点,摒弃低信誉的节点。模拟船联网网络结构进行仿真试验,分别采用4种不同的攻击方法对网络中部分节点进行攻击,试验分别按迭代100、200、300次进行。仿真结果表明:传统方法的数据融合准确率为78%,信任模型的数据融合准确率达到93%,结果更逼近真实数据,与传统方法相比有效提高了融合的准确性和可靠性。
The concepts of subjective trust elevation and its expected probability were introduced. The credibility value of every node in the internet of ships was evaluated by statistical method, and calculated lay using KL divergence. A trust model of data fusion was proposed based on node credibility evaluation. The nodes with high credibilities were included in the model, however, the nodes with low credibilities were abandoned in order to avoid their effects in data fusion. The structure of internet of ships was simulated, four different kinds of attacks were adopted to attack some network nodes, and simulation experiments were done where the iteration numbers were 100, 200 and 300 respectively. Simulation result shows that data fusion accuracy is 78% by using traditional method, and fusion result by using trust model is nearer to true data with an accuracy of 93%. Compared with traditional method, the accuracy and security of data fusion are efficiently improved by trust model. 1 tab, 8 figs, 17 refs.
出处
《交通运输工程学报》
EI
CSCD
北大核心
2013年第3期121-126,共6页
Journal of Traffic and Transportation Engineering
基金
国家自然科学基金项目(51278058)
长江学者和创新团队发展计划项目(IRT0951)
中央高校基本科研业务费专项资金项目(CHD2011JC056)
陕西省自然科学基金项目(2012JM8011)
关键词
物联网
船联网
数据融合
信任模型
信任评估
KL距离
信誉度
internet of things
internet of ships
data fusion
trust model
trust evaluation
KLdivergence
credibility
