无线传感器网络节点网格搜索定位  

Grid Scan Method for Wireless Sensor Network Node Localization

作　　者：陈杰[1] 何磊[1] 

机构地区：[1]中国科学院苏州纳米技术与纳米仿生研究所,江苏苏州215123

出　　处：《计算机测量与控制》2013年第6期1721-1722,1726,共3页Computer Measurement &Control

基　　金：苏州市科技支撑计划项目(SS201029);苏州市科技计划应用基础研究项目(SYN201119)

摘　　要：节点定位是无线传感器网络关键技术之一;首先,根据信号传播模型-对数正态阴影衰落模型,确定基于接收信号强度的最大似然目标函数。因为无法通过求导的方法求出目标方程的解析解,所以这里采用网格搜索的方法,来确定节点的最优坐标;在考虑如何进行网格搜索时,借鉴图像压缩算法中的三步搜索算法、共轭搜索算法等,来确定待测试点的坐标;当待测坐标中使目标函数值达到最大时,该点即为未知节点的最优坐标;本文仿真了上述算法,与现有的一些定位算法相比:复杂度低,无需测量节点之间的距离,精度有了10%左右的提高。Node localization is one of key technologies of wireless sensor network. First, according to Logarithmic Attenuation Model, a maximum likelihood target function based on RSSI has been founded. Because analytical solution of the target function is hard to be worked out by derivation, the coordinate of unknown node is found through grid scanning. It refers Three--Step Search Algorithm and Conjugate Scan Algorithm which are often used in image compression algorithm to get testing point coordinates. When a coordinate makes the target function reach maximum, it can be the most optimal coordinate of the unknown node. The algorithm is more accurate and is less complex, and doesn＇t need to measure distance between nodes, compared with other popular algorithms. The localization accuracy is improved about 10%.

关 键 词：无线传感器网络 节点定位 最大似然准则 网格搜索算法 

分 类 号：TP212[自动化与计算机技术—检测技术与自动化装置]