WiDir: walking direction estimation using wireless signals
WiDir: walking direction estimation using wireless signals
复制标题
DOI：
10.1145/2971648.2971658
复制
发表时间：
2016-09
期刊：
影响因子：
--
通讯作者：
Dan Wu;Daqing Zhang;Chenren Xu;Yasha Wang;Hao Wang
中科院分区：
文献类型：
--
作者：
Dan Wu;Daqing Zhang;Chenren Xu;Yasha Wang;Hao Wang
尽管它很重要，但步行方向仍然是一个关键的环境，缺乏一个经济有效的、连续的解决方案，人们可以在室内环境中访问。最近，无设备传感引起了人们的极大关注，因为这些技术不需要用户携带任何设备，因此可以在智能家居和办公室中实现许多应用。在本文中，我们介绍了WiDir，这是第一个利用WiFi无线信号以无设备方式估计人类行走方向的系统。人体运动改变了多径分布，从而改变了接收端的WiFi信道状态信息。WiDir基于菲涅耳区域模型分析多个WiFi子载波的相变动态，推断行走方向。我们使用商用WiFi设备实现了一个概念验证原型，并在家庭和办公室环境中对其进行了评估。实验结果表明，WiDir能够以小于10度的中值误差估计人类的行走方向。
Despite its importance, walking direction is still a key context lacking a cost-effective and continuous solution that people can access in indoor environments. Recently, device-free sensing has attracted great attention because these techniques do not require the user to carry any device and hence could enable many applications in smart homes and offices. In this paper, we present WiDir, the first system that leverages WiFi wireless signals to estimate a human's walking direction, in a device-free manner. Human motion changes the multipath distribution and thus WiFi Channel State Information at the receiver end. WiDir analyzes the phase change dynamics from multiple WiFi subcarriers based on Fresnel zone model and infers the walking direction. We implement a proof-of-concept prototype using commercial WiFi devices and evaluate it in both home and office environments. Experimental results show that WiDir can estimate human walking direction with a median error of less than 10 degrees.