PAMT: Phase-based Acoustic Motion Tracking in Multipath Fading Environments
PAMT: Phase-based Acoustic Motion Tracking in Multipath Fading Environments
复制标题
DOI：
10.1109/infocom.2019.8737366
复制
发表时间：
2019-04
期刊：
影响因子：
--
通讯作者：
Yang Liu;Wuxiong Zhang;Yang Yang-Yang;Weidong Fang;F. Qin;Xuewu Dai
中科院分区：
文献类型：
--
作者：
Yang Liu;Wuxiong Zhang;Yang Yang-Yang;Weidong Fang;F. Qin;Xuewu Dai
运动跟踪技术已被广泛应用于虚拟现实(VR)、健康监测和虚拟触摸控制等移动交互应用中。与专用硬件设备相比，手机使用可靠的扬声器和麦克风，可以作为基于廉价声学的运动跟踪解决方案的无处不在的设备。然而，对于复杂的室内环境，由于移动设备的多径衰落和有限的采样率，基于声学的方法很难实现准确的运动跟踪。本文定义了一个新的参数--多径效应比(MER)来表示不同频率下接收信号的多径衰落效应。基于MER，提出了一种新的多径效应抑制技术，利用多个扬声器计算声信号的相位变化并跟踪相应的运动距离。然后提出了一种基于相位的声学运动跟踪(PAMT)方法，并在标准Android智能手机上实现了该方法。实验结果表明，在不需要任何专用硬件的情况下，PAMT在多径衰落环境下的定位和运动跟踪应用中可以达到令人印象深刻的毫米级精度。具体而言，一维场景和二维场景的测量误差分别小于2 mm和4 mm。
Motion tracking technologies have been widely used in mobile interaction applications, such as Virtual Reality (VR), healthy monitoring, and virtual touch control. Compared with dedicated hardware devices, mobile phones use reliable speakers and microphones, and can serve as ubiquitous devices for cheap acoustic-based motion tracking solutions. However, for complex indoor environments, it is very difficult for acoustic-based methods to achieve accurate motion tracking due to multipath fading and limited sampling rate at mobile devices. In this paper, a new parameter named Multipath Effect Ratio (MER) is defined to indicate the multipath fading effect on received signals at different frequencies. Based on MER, a novel multipath effect mitigating technique is developed to calculate the phase change of acoustic signals and track the corresponding moving distance by using multiple speakers. A Phase-based Acoustic Motion Tracking (PAMT) method is then proposed and implemented on standard Android smartphones. Experiment results show, without any specialized hardware, PAMT can achieve an impressive millimeter-level accuracy for localization and motion tracking applications in multipath fading environments. Specifically, the measurement errors are less than 2mm and 4mm in one-dimensional and two-dimensional scenarios, respectively.