Perakis Harris, Gikas Vassilis, Retscher Günther
School of Rural, Surveying and Geoinformatics Engineering, National Technical University of Athens, 157 80 Zographos, Greece.
Department of Geodesy and Geoinformation, TU Wien-Vienna University of Technology, 1040 Vienna, Austria.
Sensors (Basel). 2024 Nov 25;24(23):7520. doi: 10.3390/s24237520.
"Smart" devices, such as contemporary smartphones and PDAs (Personal Digital Assistance), play a significant role in our daily live, be it for navigation or location-based services (LBSs). In this paper, the use of Ultra-Wide Band (UWB) and Wireless Fidelity (Wi-Fi) based on RTT (Round-Trip Time) measurements is investigated for pedestrian user localization. For this purpose, several scenarios are designed either using real observation or simulated data. In addition, the localization of user groups within a neighborhood based on collaborative navigation (CP) is investigated and analyzed. An analysis of the performance of these techniques for ranging the positioning estimation using different fusion algorithms is assessed. The methodology applied for CP leverages the hybrid nature of the range measurements obtained by UWB and Wi-Fi RTT systems. The proposed approach stands out due to its originality in two main aspects: (1) it focuses on developing and evaluating suitable models for correcting range errors in RF-based TWR (Two-Way Ranging) technologies, and (2) it emphasizes the development of a robust CP engine for groups of pedestrians. The results obtained demonstrate that a performance improvement with respect to position trueness for UWB and Wi-Fi RTT cases of the order of 74% and 54%, respectively, is achieved due to the integration of these techniques. The proposed localization algorithm based on a P2I/P2P (Peer-to-Infrastructure/Peer-to-Peer) configuration provides a potential improvement in position trueness up to 10% for continuous anchor availability, i.e., UWB known nodes or Wi-Fi access points (APs). Its full potential is evident for short-duration events of complete anchor loss (P2P-only), where an improvement of up to 53% in position trueness is achieved. Overall, the performance metrics estimated based on the extensive evaluation campaigns demonstrate the effectiveness of the proposed methodologies.
“智能”设备，如当代的智能手机和个人数字助理（PDA），在我们的日常生活中发挥着重要作用，无论是用于导航还是基于位置的服务（LBS）。在本文中，研究了基于往返时间（RTT）测量的超宽带（UWB）和无线保真（Wi-Fi）用于行人用户定位的情况。为此，设计了几种场景，使用实际观测数据或模拟数据。此外，还研究和分析了基于协作导航（CP）的邻域内用户群体的定位。评估了使用不同融合算法对这些用于测距定位估计技术的性能分析。应用于CP的方法利用了UWB和Wi-Fi RTT系统获得的距离测量的混合特性。所提出的方法在两个主要方面因其创新性而脱颖而出：（1）它专注于开发和评估用于校正基于射频的双向测距（TWR）技术中距离误差的合适模型，以及（2）它强调为行人群体开发强大的CP引擎。获得的结果表明，由于这些技术的集成，UWB和Wi-Fi RTT情况下的位置真实性分别提高了约74%和54%。所提出的基于对等基础设施/对等（P2I/P2P）配置的定位算法在连续锚点可用性（即UWB已知节点或Wi-Fi接入点（AP））的情况下，位置真实性有高达10%的潜在提高。对于完全锚点丢失的短持续时间事件（仅P2P），其全部潜力很明显，此时位置真实性提高了高达53%。总体而言，基于广泛评估活动估计的性能指标证明了所提出方法的有效性。
