Trajectory identification based on spatio-temporal proximity patterns between mobile phones
Trajectory identification based on spatio-temporal proximity patterns between mobile phones
复制标题
基于手机间时空邻近模式的轨迹识别
DOI：
10.1007/s11276-015-0987-z
复制
发表时间：
2016
影响因子：
3
通讯作者：
Takamasa Higuchi，Hirozumi Yamaguchi，Teruo Higashino
中科院分区：
文献类型：
--
作者：
Tsuchiya;H.;Burana;D.;Ohtake;F.;Arai;N.;Kaiho;A.;Komada;M.;Tanaka,K.;and Saeki;Y.;Takamasa Higuchi，Hirozumi Yamaguchi，Teruo Higashino
位置相关的移动的应用的日益普及不断刺激对定位技术的需求。然而，尽管在过去的十年中的重大研究工作，在室内环境中的精确定位仍然是一个悬而未决的问题。在本文中，我们提出了一种新型的室内定位系统，该系统通过有效地利用激光测距扫描仪的强大行人跟踪能力（即，基于激光的距离测量传感器）。虽然基于激光的跟踪系统可以精确地检测在其感测区域中的每个位置处行人的存在，但是位置信息不与任何移动的电话用户相关联，因此其基本上不能提供用户自己的位置。为了消除这种限制，我们关注移动的电话之间的时空邻近模式，其可以通过对等短距离无线通信（例如，蓝牙）。通过检查通信日志和基于激光的跟踪检测到的anonymousjectories之间的接近度之间的一致性，我们的系统识别出对应于每个移动的手机用户的轨迹，以提供他们自己的位置信息。通过大量的仿真和现场实验，我们表明，我们的系统可以实现高达91%的轨迹识别精度。
Growing popularity of location-dependent mobile applications is continuously stimulating a demand for localization technology. However, in spite of significant research effort in the past decade, precise positioning in indoor environments is still an open problem. In this paper, we propose a novel type of indoor localization system that provides mobile phone users in a pedestrian crowd with theirownposition information of sub-meter accuracy by effectively utilizing a powerful pedestrian tracking capability of laser range scanners (i.e., laser-based distance measurement sensors). Although the laser-based tracking system can precisely detect presence of pedestrians at each location in its sensing region, the location information is not associated with any mobile phone users and thus it basically cannot provide the users’ own locations. To remove this limitation, we focus on spatio-temporal proximity patterns between mobile phones, which can be detected by peer-to-peer short-range wireless communication (e.g., Bluetooth). By examining consistency between the communication logs and proximity betweenanonymoustrajectories detected by laser-based tracking, our system identifies a trajectory that corresponds to each mobile phone user to offer their own position information. Through extensive simulations and field experiments, we show that our system can achieve trajectory identification accuracy of up to 91 %.