基于结构化噪声矩阵补全的WSNs收集数据重建方法

发布时间：2018-04-17 16:28

本文选题：无线传感器网络 + 数据收集　；参考：《数据采集与处理》2017年05期

【摘要】：许多科学研究都需要对环境数据进行分析,这些环境数据通常是通过部署在研究区域内的无线传感器网络(Wireless sensor networks,WSNs)来收集的。收集数据的完整性和准确性决定了科研结果的可靠性。然而,在数据收集过程中普遍存在的数据丢失和错误影响了收集数据的可用性,为此需要利用收集到的数据重建完整的环境数据。基于环境数据低秩特性,将数据重建问题建模为L2,1范数正则化矩阵补全模型,提出一种基于结构化噪声矩阵补全的WSNs收集数据重建方法(Data reconstruction approach via matrix completion with structural noise,DRMCSN)。真实数据集上的实验结果表明,该方法性能优于现有算法,不仅能以较高的精度恢复缺失的环境数据,而且能辨识出收集到错误数据的传感器节点。
[Abstract]:Many scientific studies require the analysis of environmental data, which is usually collected through Wireless sensor Networks (WSNs) deployed in the study area.The completeness and accuracy of collecting data determine the reliability of scientific research results.However, the data loss and error in the process of data collection affect the availability of data collection, so we need to use the collected data to reconstruct the complete environment data.Based on the low rank characteristic of environment data, the data reconstruction problem is modeled as L2G norm regularization matrix complement model, and a WSNs data collection method based on structured noise matrix complement is proposed. The data reconstruction approach via matrix completion with structural noiseDRCSNs are presented.The experimental results on real data sets show that the proposed method is superior to the existing algorithms and can not only recover the missing environmental data with high accuracy, but also identify the sensor nodes that collect the wrong data.
【作者单位】：金陵科技学院电子信息工程学院;江苏省无线传感网高技术研究重点实验室;南京邮电大学计算机学院;
【基金】：江苏省自然科学基金(BK20130096,BK20161516,BK20161104)资助项目国家自然科学基金(61300240,61572263)资助项目江苏省高校自然科学基金(15KJB520027)资助项目中国博士后科学基金(2015M581794)资助项目江苏省博士后科研资助计划(1501023C)资助项目安徽省自然科学基金(1608085MF127)资助项目金陵科技学院高层次人才工作启动(JIT-201527)资助项目
【分类号】：TN929.5;TP212.9

【相似文献】