Abstract:Recently, with the development of automatic driving, high-precision localization has become indispensable for safety. The inertial measurement unit (IMU) integrated with ultra-wideband (UWB) technology has been thought as a promising scheme by the company Humatics for trains. Generally, the spaces of transportation are extended along a long path with flexural tunnels and sideway obstacles around, where non-line-of-sight (NLoS) propagation happens frequently. However, a large interval deployment of UWB anchors along such narrow path with NLoS interference will dramatically decrease localization performance. Hence, we propose a robust algorithm to mitigate the NLoS interference. Firstly, a joint LoS/NLoS detection and mitigation algorithm is proposed to improve the ranging accuracy of mobile target with UWB tags under mixed LoS/NLoS interference. Secondly, we improve the conventional Kalman Filter (KF) algorithm with forward-backward propagation to integrate temporal correlations further. Finally, a comprehensive fusion localization scheme with ranging error mitigation and improved KF is proposed. The simulation results show that, regarding the static localization, the proposed joint LoS/NLoS detection and mitigation algorithm outperforms five other typical NLoS elimination algorithms, and achieves 48.2%, 62.4%, 45.5%, 69.9% and 50.2% gains in terms of root mean square error (RMSE), under NLoS scenarios with a mean ranging error of 0.5 m. Regarding the dynamic localization, compared with two other typical KF localization algorithms, the proposed localization scheme achieves 20.9% and 14.5% localization accuracy improvements in LoS scenarios, 45.7% and 47.2% improvements in mixed LoS/NLoS scenarios, respectively. Furthermore, the effectiveness of the proposed scheme is also verified in our testing platform.

Mobi-Track: Distilling Direct Path In Time Domain For High Accuracy Wifi Tracking

Signal Tracking Using Commodity WiFi

LTrack: A LoRa-Based Indoor Tracking System for Mobile Robots

Fast-Tracker 2.0: Improving Autonomy of Aerial Tracking with Active Vision and Human Location Regression.

PAssTrack: Practical and Accurate Passive Human Tracking System Using Commodity Wi-Fi

A Localization and Tracking System Using Single WiFi Link

STrack: Velocity Estimation Using Single Antenna WiFi Devices

Toward Reliable Non-Line-of-Sight Localization Using Multipath Reflections

IndoTrack

Single-Target Real-Time Passive WiFi Tracking

WiTraj: Robust Indoor Motion Tracking With WiFi Signals

Online Trajectory Estimation Based on a Network-Wide Cellular Fingerprint Map

TraIL: Pinpoint Trajectory for Indoor Localization

Tracking from One Side - Multi-Person Passive Tracking with WiFi Magnitude Measurements

Widar: Decimeter-Level Passive Tracking via Velocity Monitoring with Commodity Wi-Fi

Decimeter Level Passive Tracking with Wifi.

Itracking: Accurate Light-Based Location-Tracking in Wireless Sensor Networks

MobiFi - Fast Deep-Learning Based Localization Using Mobile WiFi.

RobMOT: Robust 3D Multi-Object Tracking by Observational Noise and State Estimation Drift Mitigation on LiDAR PointCloud

TriLoc: Toward Accurate Indoor Localization With Assistance of Microwave Reflections

Robust Localization for Mobile Targets Along a Narrow Path With LoS/NLoS Interference