“ViPER: Vehicle Pose Estimation using Ultra-WideBand Radios” by Alireza Ansaripour, Milad Heydariaan, and Omprakash Gnawali. In Proceedings of the annual International Conference on Distributed Computing in Sensor Systems (DCOSS 2020), June 2020.
Pose estimation is a building block for many location-based applications, such as safety applications in a construction site. Ultra-WideBand (UWB) Radios have been widely used for localization and can be used in pose (location and orientation angle of the object) estimation primarily because of the accuracy with which these radios can estimate the arrival time of radio signals. Current UWB pose estimation solutions do not perform adequately in Non-Line of Sight (NLoS) conditions. Some of these existing solutions in pose estimation rely on two or more types of sensors to tackle the NLoS challenge. These methods suffer from data fusion complexity, making the system not generalizable and limited to some specific simple environments, such as labs. In this paper, we propose ViPER, a UWB-based pose estimating system using only UWB radios. Our goal is to reduce the effects of the NLoS without the inclusion of any auxiliary sensors. ViPER uses low-pass filter, anchor and reference selection method to reduce the effect of NLoS in the measurements. It also estimates the pose of the entities using an optimization problem. We have evaluated ViPER in real-world highway construction and parking lot setting. We find that it improves the average packet reception ratio by 117% and decreases the error rate by 70% over the state of the art in Non-Line of Sight situation.
Download Presentation Slides.BibTeX entry:
@inproceedings{ansaripour2020viper,
title={ViPER: Vehicle Pose Estimation using Ultra-WideBand Radios},
author={Ansaripour, Alireza and Heydariaan, Milad and Gnawali, Omprakash},
booktitle={2020 16th International Conference on Distributed Computing in Sensor Systems (DCOSS)},
year={2020},
organization={IEEE}
}