Indoor Localization within Multi-Story Buildings Using MAC and RSSI Fingerprint Vectors
Indoor Localization within Multi-Story Buildings Using MAC and RSSI Fingerprint Vectors
复制标题
使用 MAC 和 RSSI 指纹向量进行多层建筑的室内定位
DOI：
10.3390/s19112433
复制
发表时间：
2019-06-01
期刊：
影响因子：
3.9
通讯作者：
Song, Shiming
中科院分区：
文献类型：
--
作者：
Han, Litao;Jiang, Li;Song, Shiming
为了使现有的无线网络设备和智能手机容易达到可用的定位精度，指纹定位被广泛应用于室内定位，其依赖于来自不同地方的无线局域网（WLAN）的接收信号强度指示符（RSSI）的差异。目前，大多数研究者关注的是利用RSSI值对在线定位算法的改进，而很少有人关注从WLAN接收到的MAC（媒体访问控制）地址。因此，我们尝试将MAC地址和RSSI值同时整合，以实现多层建筑物内的室内定位。本文提出了一种多层建筑室内定位的新方法，该方法包括两个步骤：首先利用不同楼层接收到的MAC地址的差异来确定楼层;然后在同一楼层进行进一步定位。同时，在在线定位阶段引入以MAC地址为聚类索引的聚类操作，提高了室内定位的效率和精度。实验结果表明，该方法不仅可以实现精确定位的水平精度为1.8米，而且可以在多层建筑物内的接收机所在的楼层。
For existing wireless network devices and smart phones to achieve available positioning accuracy easily, fingerprint localization is widely used in indoor positioning, which depends on the differences of the Received Signal Strength Indicator (RSSI) from the Wireless Local Area Network (WLAN) in different places. Currently, most researchers pay more attention to the improvement of online positioning algorithms using RSSI values, while few focus on the MAC (media access control) addresses received from the WLAN. Accordingly, we attempt to integrate MAC addresses and RSSI values simultaneously in order to realize indoor localization within multi-story buildings. A novel approach to indoor positioning within multi-story buildings is presented in this article, which includes two steps: firstly, to identify the floor using the difference of received MAC addresses in different floors; secondly, to implement further localization on the same floor. Meanwhile, clustering operation using MAC addresses as the clustering index is introduced in the online positioning phase to improve the efficiency and accuracy of indoor positioning. Experimental results show that the proposed approach can achieve not only the precise location with the horizontal accuracy of 1.8 meters, but also the floor where the receiver is located within multi-story buildings.