基于证据推理规则的无线传感器网络数据可靠性分析

Data Reliability Analysis of Wireless Sensor Networks Based on Evidence Reasoning Rules

无线传感器网络(Wireless Sensor Network,WSN)在精密工程中具有广泛的应用,对数据精度的要求严苛,因此对WSN可靠性进行评估具有实用价值。本文以WSN实际监测数据为研究对象,通过对WSN数据提取特征,选取所测环境数据的时间相关性和空间相关性以及节点电压为可靠性指标,提出了基于证据推理规则(Evidential reasoning rule,ER)的WSN数据可靠性评估模型。该模型采用变异系数法和基于距离的方法,确定评估指标权重和可靠度,以基于规则的方法将指标数据统一成置信分布形式,利用证据推理规则对指标和参数进行融合,得到WSN数据可靠性状态。最后,通过实例分析验证该模型的有效性。

In practical engineering applications,stringent requirements are presented for the reliability of collected information under the operative mode of wireless sensor network(WSN),therefore,assessing the reliability of WSN and guaranteeing that WSN is at a healthy operative mode have a practical value.In the present stage,the researches on the reliability of WSN mainly focus on various transmission protocols and data processing algorithms.In this paper,the actual monitoring data of WSN are taken as the research object,the reliability evaluation model of WSN data based on evidential reasoning rule is put forward by extracting features of WSN,selecting temporal correlation,spatial correlation and node voltage of measured environmental data as reliability indexes.In this model,the variation coefficient method and the distance-based method are adopted,the weight and reliability of fusion evidence are determined,the index data are unified into a form of confidence distribution with the rule-based method,the indexes and parameters are integrated through evidential reasoning rule,so as to obtain the reliability status of WSN.Finally,the effectiveness of this model is verified through instance analysis.