Range-Based Localization Algorithm for Next Generation Wireless Networks Using Radical Centers
Range-Based Localization Algorithm for Next Generation Wireless Networks Using Radical Centers
复制标题
DOI：
10.1109/access.2016.2551704
复制
发表时间：
2016
期刊：
影响因子：
3.9
通讯作者：
Yeong-Sheng Chen;Der-Jiunn Deng;Chun-Chien Teng
中科院分区：
文献类型：
--
作者：
Yeong-Sheng Chen;Der-Jiunn Deng;Chun-Chien Teng
定位是下一代无线网络（NGWN）（诸如5G蜂窝网络、IEEE 802.11ax WiFi网络和无线自组织传感器网络（WASN））中的重要问题。本文利用解析几何中的根中心概念，提出了一种下一代网络的三维定位算法，并以WASNs为例进行了研究。假设未知节点可以测量其到四个或更多个锚节点（参考节点）的距离。通过拾取到四个锚节点的四个距离测量，计算径向中心。通过解析公式，径向中心示出能够被视为未知节点位置的估计。由于每四个距离测量产生一个径向中心（在3-D空间中），为了进一步改进和融合这些估计，提出了有效的过滤机制来过滤掉不适当的估计。然后，剩余的径向中心被平均，并且解是未知节点位置的最终估计。分析比较了该算法与传统最小均方误差（MMSE）方法的定位误差。仿真结果表明，该算法在精度和效率上均优于传统的MMSE方法。大量的计算机仿真进行了验证，结果表明，所提出的定位算法的优势比MMSE方法。
Localization is an important issue in next generation wireless networks (NGWNs), such as 5G cellular networks, IEEE 802.11ax WiFi networks, and wireless ad hoc sensor network (WASNs). This paper proposes a 3-D localization algorithm for NGWNs by utilizing the concept of radical centers from analytic geometry, and takes WASNs as an example. Assume that an unknown node can measure its distances to four or more anchor nodes (reference nodes). By picking four distance measurements to four anchor nodes, a radical center is computed. Through analytic formulation, the radical center is shown to be able to be treated as an estimation of the unknown node location. Since every four distance measurements generates one radical center (in 3-D space), to further improve and fuse these estimations, effective filtering mechanisms are proposed to filter out the improper estimations. Afterward, the remaining radical centers are averaged, and the solution is the final estimation of the unknown node location. The location errors of the proposed algorithm and the conventional minimum mean square error (MMSE) method were analytically compared. It was shown that the proposed algorithm outperforms the conventional MMSE method both in accuracy and efficiency. Extensive computer simulations were carried out and the results verified the advantage of the proposed location algorithm over the MMSE approach.