无线传感器网络簇内自适应融合算法研究  

Study of adaptive data fusion algorithm within clusters for WSN

作　　者：马琦[1] 张记龙[1,2] 王志斌[1,2] 

机构地区：[1]中北大学山西省光电信息与仪器工程技术研究中心,太原030051 [2]中北大学仪器科学与动态测试教育部重点实验室,太原030051

出　　处：《计算机应用研究》2009年第7期2502-2503,2507,共3页Application Research of Computers

基　　金：国家科技部国际科技合作重点资助项目(2006DFB72510);山西省国际合作资助项目(2006081015)

摘　　要：无线传感器网络中采集的数据存在着较大的冗余与误差,且影响数据的可靠性。针对这个问题,分析了簇内数据误差成因,提出了一种改进后的自适应数据融合算法。该算法从节点测量数据自身着手,通过迭代得到各个节点测量数据的无偏估计值,以各个节点与估计值的欧氏距离作为各节点可信度的描述。实验证明,该融合算法提高了数据的精度和可信度。同时,通过与分批估计融合方法和传统的自适应加权融合方法的比较分析,表明该方法融合效果更好。There are many redundances and errors in the data acquired from wireless sensor network, which has bad effect in the dependability of the data. Aiming at this problem, this paper analyzed the reason of the errors in WSN, and presented an improved adaptive data fusion algorithm. This algorithm got the unbiased estimates of every node through iterative methods based on the node testing data. It made the Euclidean distance between every node and estimates as the precision of the relia- bility. Experiment results show that the fusion algorithm enhance the precision and reliability of the data. Compared with the batch estimation fusion method and traditional adaptive weighting fusion method, this fusion algorithm has better fusion perforrnance.

关 键 词：无线传感器网络 数据融合 自适应加权 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]