Yin Lu, Ni Qiang, Deng Zhongliang
IEEE Trans Cybern. 2021 Apr;51(4):2188-2200. doi: 10.1109/TCYB.2019.2900312. Epub 2021 Mar 17.
Localization plays a key role in Internet of Things. This paper proposes a novel intelligent cooperative multisensor localization method called the edge cloud cooperative localization (ECCL) which has the range and angle observations from the neighbor nodes along with the location observations from an absolute coordinate localization system like global positioning system. The edge cloud structure is proposed which employs several distributed Kalman filters in sensor nodes edge and a centralized cooperative fusion unit in the cloud. For a robust fusion, a cooperative redundancy validation method is proposed to detect the outliers. The proposed ECCL scheme has the advantages of both the distributed and centralized localization, which satisfies the needs of high reliability and high accuracy, especially when sensor nodes have limited computational resources. The simulation and experimental results show that our proposed ECCL algorithm outperforms the other schemes both in outlier detection and localization accuracy.
定位在物联网中起着关键作用。本文提出了一种新颖的智能协作多传感器定位方法，称为边缘云协作定位（ECCL），该方法具有来自相邻节点的距离和角度观测值，以及来自诸如全球定位系统等绝对坐标定位系统的位置观测值。提出了边缘云结构，该结构在传感器节点边缘采用多个分布式卡尔曼滤波器，并在云端采用一个集中式协作融合单元。为了进行稳健融合，提出了一种协作冗余验证方法来检测异常值。所提出的ECCL方案具有分布式和集中式定位的优点，满足了高可靠性和高精度的需求，特别是在传感器节点计算资源有限的情况下。仿真和实验结果表明，我们提出的ECCL算法在异常值检测和定位精度方面均优于其他方案。
