周界监测中的阈值优化和多模态节点协同检测算法

A New Optimization Method for Thresholds and Collaborative Targets Detection in Perimeter Monitoring

为提高无线传感器网络周界监测中多节点的目标协同检测能力,提出了一种新的阈值优化方法及多模态节点协同检测算法。首先,在一定的环境噪声下,采用蒙特卡罗方法建立节点虚警率与节点阈值之间、系统虚警率与系统阈值和节点虚警率之间的映射关系,通过查表直接获取阈值,实现了节点和系统阈值优化;然后,利用节点的空间分布特性和目标信号信噪比,对节点检测结果进行加权修正,实现了多模态节点的检测结果融合;最后,通过仿真实验对阈值优化方法和多模态节点协同检测算法进行了验证。仿真结果显示,相对于单模态节点协同检测算法和简单阈值判决算法,多模态节点协同检测算法的目标检测率分别提高了约25%和3%。

A new optimization method for thresholds and a multi-modal nodes collaborative detecting algorithm are proposed to improve the target detection capability of perimeter monitoring in wireless sensor networks.The optimization of the node and the system thresholds is realized by using the Monte Carlo method to create the mapping table between the node false alarm rates and the node thresholds,as well as the mapping table among the system false alarm rates,the node false alarm rates and system thresholds.Then,the optimal node and system thresholds are obtained by looking up the mapping tables.The results of the multi-modal nodes collaborative detecting is fuzed by using the distributed characteristics of nodes and target signal-to-noise ratio to revise the weighting of the node detecting results.Simulation experiments and comparisons with the single-modal node collaborative detecting algorithm and the simple thresholds detecting algorithm show that the multi-modal nodes collaborative detecting algorithm increases the target detection rate by about 25% and 3%,respectively.