摘要： 针对无线传感器网中节点无标识以及数量未知环境下的节点定位问题,提出一种通过优化圆环交叉区域筛选可行节点位置和数量的算法。采用粗粒度的圆环搜索标识重叠区域的交叉数量,生成二阶定位点权重矩阵。通过求矩阵极大值确定并筛选出可能含有未知节点的圆环交叉区域,利用每个区域的质心代表该交叉区域。运用自适应遗传算法估计未知节点的数量和位置,将贝叶斯信息准则最小值作为选择模型参数最优值的依据。实验结果表明,在未知节点分布稀疏的情况下,该定位算法既能准确估计出未知节点的数量,也能达到较高的定位精度。

关键词: 无线传感器网络, 定位, 节点无标识, 圆环搜索, 自适应遗传算法

Abstract: In Wireless Sensor Network(WSN) which consists of node without identification and quantity,this paper proposes an algorithm in order to solve the problem of node localization.This algorithm can achieve nodes’ quantity and locations by optimizing the ring crossing area.The algorithm generates position weight matrix by employing coarse-grained ring search and identifying cross-quantity of the overlapping area.It identifies and filters out the ring crossing area which may contain unknown nodes by calculating the maximum value,and on behalf of the intersection area using the centroid of each region.It makes use of adaptive genetic algorithm to estimate the quantity and locations of unknown nodes,and takes the minimum value of Bayesian Information Criterion(BIC) as the basis for choosing the optimal parameters of the model parameters.Experimental results show that the algorithm can accurately estimate the quantity of unknown nodes,and the algorithm can achieve a higher positioning accuracy in the case of sparse distribution of the unknown nodes.

Key words: Wireless Sensor Network(WSN), location, node without identification, ring search, adaptive genetic algorithm

中图分类号: 

• TP393