基于卡方拟合度的无线传感器网络数据复原汇聚方法

摘要

在无线传感器网络（ WSNs）中，现有的数据复原汇聚算法不能准确判断节点感知数据的受攻击程度，数据复原精度偏低，故提出了一种基于卡方拟合度的分布式数据复原汇聚算法。该算法根据不同时刻节点感知数据的时间相关性特点来构造各节点信任权值计算当前时刻各簇数据样本的估计值，并利用卡方拟合度来衡量此时各个簇的受攻击程度，最后通过加权运算提高了算法的数据复原精度。另外，由于卡方拟合度能够准确感知数据的细微波动，该方法对网络噪声干扰的稳健性有很大提高。仿真结果表明：该算法的数据复原汇聚精度大幅度提高，优于现有数据复原汇聚算法。%The existing resilient data aggregation algorithm doesn’t accurately judge under-attack level of each node in wireless sensor networks( WSNs ),resulting in a low precision,propose a distributed resilient data aggregation algorithm based on Chi-square goodness of fitting. The algorithm firstly construct weights of trust of each node,calculates current estimation value of data sample of each cluster by exploiting time correlation of a sensor node at different time,to evaluate attack level using Chi-square goodness of fitting. Finally,through weighted arithmetic,increase precision of data recovery. In addition,Chi-square goodness of fitting can sense to slight data fluctuations,accurately thus the presented algorithm can better improve robust of network noise. Simulation result shows that data aggregation precision of the presented algorithm is greatly improved,which is better than existing data resilient aggregation algorithm.