...
  • 主题
高级搜索  >
历史搜索 清空历史
首页> 外文期刊>Computer Communications >Efficient target detection in maritime search and rescue wireless sensor network using data fusion
【24h】

Efficient target detection in maritime search and rescue wireless sensor network using data fusion

机译：使用数据融合的海上搜索与救援无线传感器网络中的有效目标检测

获取原文
获取原文并翻译 | 示例
   

获取外文期刊封面封底 >>

       
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

Maritime search and rescue wireless sensor network (MSR-WSN) has been a bedrock to discover the floating target after the shipwreck. In this paper, we first define a sea region of target detection and formulate a clustered topology of MSR-WSN. Second, we employ the sensor nodes of MSR-WSN to track the collective radio signal emitted by the mobile target. Each node firstly transmits the preprocessed perceived data to the cluster head node. Next, the data fusion center (DFC) collects a local decision of cluster head node through a binary hypothesis test and works out an accurate global decision. This paper emphasizes at designing both local and global data fusion rules based on the likelihood of ratio test statistics using a Neyman-Pearson lemma and Bayesian approach. One major stumbling block in the ocean lies in a complex and changing communication environment. There is a need for the DFC to develop a fusion rule of carrying out a dependable target detection to screen out the side effect of wave shadow. To address the concern, we propose a novel mobile target detection algorithm (NMTDA) based on information theory. The main idea is to dynamically calculate an adaptive decision threshold using both Kullback-Leibler divergence (KLD) and a global optimal decision statistics to enforce the accuracy of target detection. In addition, KLD is adopted to quantify the strength of wave shadow effect and tune Correct Detection/Flase Alarm probabilities of target detection. To conserve the overall MSR-WSN energy, DFC selects clusters with the maximum predictive information gain for MSR before next round search. Extensive simulation results demonstrate that our proposed mobile target detection algorithm works well in maritime search and rescue scenario.
机译：海上搜索和救援无线传感器网络（MSR-WSN）已成为海难后发现漂浮目标的基石。在本文中，我们首先定义了目标检测的海域，并制定了MSR-WSN的聚类拓扑。其次，我们使用MSR-WSN的传感器节点来跟踪移动目标发射的集合无线电信号。每个节点首先将预处理后的感知数据传输到集群头节点。接下来，数据融合中心（DFC）通过二元假设检验收集群集头节点的本地决策，并得出准确的全局决策。本文着重在利用Neyman-Pearson引理和贝叶斯方法基于比率检验统计数据的可能性设计局部和全局数据融合规则。海洋中一个主要的绊脚石在于复杂而多变的通讯环境。 DFC有必要制定一种融合规则，以执行可靠的目标检测来屏蔽波影的副作用。为了解决这一问题，我们提出了一种基于信息论的新型移动目标检测算法（NMTDA）。主要思想是使用Kullback-Leibler散度（KLD）和全局最佳决策统计信息动态计算自适应决策阈值，以增强目标检测的准确性。此外，采用KLD量化波影效应的强度并调整目标检测的“正确检测/火焰警报”概率。为了节省总体MSR-WSN能量，DFC在下一轮搜索之前选择MSR具有最大预测信息增益的聚类。大量的仿真结果表明，我们提出的移动目标检测算法在海上搜索和救援场景中效果很好。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号