摘要
为了减少无线传感器网络(WSN)节点在传统方法上的定位误差,增强定位的准确度,提出了一种融合改进的麻雀搜索算法和广义回归神经网络(ISSA-GRNN)的节点定位优化算法。首先,对普通DV-Hop算法和Centroid定位算法的节点信息分别优化,利用加权优化思想和节点信号强度修正DV-Hop算法的跳距与Centroid算法的质心。然后将修正后的跳距、质心特征和节点其他特征相融合,作为GRNN的输入向量进行训练。为了解决网络调节参数随机设置的问题,通过ISSA改进网络参数,并得到未知节点的最优预测位置。仿真结果表明,与其他优化算法相比,该算法平均定位误差较小,定位精度得以提升。
A node localization optimization algorithm combining improved sparrow search algorithm and generalized regression neural network(ISSA-GRNN)is proposed so as to reduce the localization errors of wireless sensor network(WSN)nodes in traditional methods and enhance the localization accuracy.Firstly,the node information of ordinary DV-Hop algorithm and Centroid positioning algorithm is optimized respectively,and the hop distance of DV-Hop algorithm and centroid of Centroid algorithm are modified by using weighted optimization method and node signal strength.Then,the modified hop distance feature,centroid feature and other features of the node are fused as the input vector of GRNN for training.As for the random setting of network adjustment parameters,the network parameters are improved by ISSA and the optimal prediction positions of unknown nodes are obtained.The simulation results show that compared with other optimization algorithms,the average localization error of this algorithm is smaller,which makes the localization accuracy greatly enhanced.
作者
王家威
薛亚辉
魏子尧
WANG Jia-wei;XUE Ya-hui;WEI Zi-yao(School of Information Engineering,Jiaozuo University,Jiaozuo 454000,China;School of Control Science and Engineering,China University of Petroleum(East China),Qingdao 266580,China)
出处
《齐鲁工业大学学报》
CAS
2022年第6期21-27,共7页
Journal of Qilu University of Technology
关键词
无线传感器网络
节点定位
加权优化
广义回归神经网络
麻雀搜索算法
wireless sensor network(WSN)
node localization
weighted optimization
generalized regression neural network(GRNN)
sparrow search algorithm(SSA)
作者简介
王家威,硕士、助教,研究方向:数据分析与处理、无线传感器网络,729687583@qq.com。
