แนวคิดหลัก
The author presents DynaPose, a system that implements dynamic anchor selection and real-time pose prediction for UWB technology, improving accuracy and efficiency in proximity services.
บทคัดย่อ
The content discusses the implementation of DynaPose, a system that enhances UWB technology for proximity services. It covers the challenges faced in UWB-based systems, the proposed solutions, experimental setups, performance evaluations, and comparisons with legacy systems.
The article introduces the concept of Ultra-wideband (UWB) technology for proximity services like UWB tagless gate (UTG). It highlights the importance of accurate localization using DL-TDoA and DS-TWR methods. The main challenge addressed is ensuring precise location determination without requiring real-time mobile device tapping.
DynaPose is introduced as a solution involving dynamic anchor selection and pose prediction based on LOS/NLOS classification using deep learning models. The system aims to improve accuracy in localization and pose prediction for MDs in real-time scenarios. The proposed system is implemented on commercial smartphones equipped with UWB modules.
Experimental setups are detailed, including datasets used for training CNN models for LOS/NLOS classification and IMU data collection for pose prediction. Results show significant improvements in accuracy compared to legacy systems, especially in NLOS conditions where DynaPose outperforms traditional methods.
Overall, the article provides insights into the advancements made in UWB technology through DynaPose implementation, showcasing improved performance in real-world scenarios.
สถิติ
DynaPose achieves a LOS/NLOS classification accuracy of 0.984.
62% higher DL-TDoA localization accuracy is achieved.
Four different poses are detected with an accuracy of 0.961 in real-time.
คำพูด
"We propose dynamic anchor selection for extremely accurate DL-TDoA localization and pose prediction for DS-TWR."
"DynaPose achieves a LOS/NLOS classification accuracy of 0.984."