cliansang / positioning-algorithms-for-uwb-matlab

The Matlab scripts for five positioning algorithms regarding UWB localization. The five algorithms are Extended Kalman Filter (EKF), Unscented Kalman Filter (UKF), Taylor Series-based location estimation, Trilateration, and Multilateration methods.

Positioning Algorithms for UWB Localization in Matlab

The Matlab scripts and its corresponding experimental data for five positioning algorithms regarding UWB localization system are provided in this repository. The five algorithms are Extended Kalman Filter (EKF), Unscented Kalman Filter (UKF), Taylor Series-based location estimation technique, Trilateration, and Multilateration methods. The UWB system under investigation is assumed as the state-space model. A constant velocity (CV) motion model and constant acceleration (CA) motion model are usable as the state model for the implementation of the mentioned UWB positioning and navigation system in this repo. In contrast to typical simulation approach for comparative analysis in literature, we use the experimental data directly achieved from the UWB hardware module manufacutured by Qorvo (Decawave) called TREK1000/EVK1000 evaluation boards.

The detailed description of the evaluated five algorithms, their implementation processes, and comparative results were addressed in Chapter 4 of my Dissertation, which is self-contained and independently readable. However, the mentioned chapter is recommended to read in conjunction with Chapter 2. A short version of the documentation is also available on our IEEE workshop paper, of which the accepted version is publicly provided in here.

The code snippets regarding the point clould transformation and registration of the Vicon camera system used in this repo were written by Timo.

Dependencies

The Control System Toolbox from mathworks was used for the implementation of UKF and EKF in this repo.

Publications

If you find this work useful in the academic context, please consider to cite one of the following papers:

@INPROCEEDINGS{Sang2019Comparative,
    author={Sang, Cung Lian and Adams, Michael and Hesse, Marc and Hörmann, Timm and Korthals, Timo and Rückert, Ulrich},
    booktitle={2019 16th Workshop on Positioning, Navigation and Communications (WPNC)}, 
    title={A Comparative Study of UWB-based True-Range Positioning Algorithms using Experimental Data}, 
    year={2019},        
    pages={1-6},
    doi={10.1109/WPNC47567.2019.8970249}
}

@INPROCEEDINGS{Sang2019Bidirectional,
    author={Sang, Cung Lian and Adams, Michael and Korthals, Timo and Hörmann, Timm and Hesse, Marc and Rückert, Ulrich},
    booktitle={2019 International Conference on Indoor Positioning and Indoor Navigation (IPIN)}, 
    title={A Bidirectional Object Tracking and Navigation System using a True-Range Multilateration Method}, 
    year={2019},        
    pages={1-8},
    doi={10.1109/IPIN.2019.8911811}
}

@ARTICLE{Sang2023Bidirectional,
    author={Sang, Cung Lian and Adams, Michael and Hesse, Marc and Rückert, Ulrich},
    journal={IEEE Journal of Indoor and Seamless Positioning and Navigation}, 
    title={Bidirectional UWB Localization: A Review on an Elastic Positioning Scheme for GNSS-deprived Zones}, 
    year={2023},
    volume={1},
    pages={161-179},
    doi={10.1109/JISPIN.2023.3337055}
} 

positioning-algorithms-for-uwb-matlab's Issues

compare_EKF_UKF_Multilat_TS_Trilat_demo_SportHall_40x20

Hello,May I ask a question?, that is, in the algorithm I asked for help (compare_EKF_UKF_Multilat_TS_Trilat_demo_SportHall_40x20) is it only the (TRILATERATION ALGORITHM USING MEASURED RANGES) algorithm that utilizes only the three flight distances A0,A1,A2; the other four algorithms utilize the four flight distances i.e. A0,A1,A1,A2 when solving for the coordinates. ) This algorithm utilizes only the A0,A1,A2 flight distances; do the other four algorithms utilize the four A0,A1,A2,A3 flight distances when solving for the coordinates? Because my current A3 coordinate measurements are not very accurate, this is what is bothering me at the moment. Looking forward to your reply, best wishes!

I have a question

Hello, I'm reading the material usefully. Excuse me, but I'd like to ask you a question, is the matlab code you uploaded possible to be implemented only with software without additional hardware? I'm asking because there is a mention of the Vicon camera code.

XY-coordinate as the input

Hi there,

I want to filter out the reading from UWB (particularly the UKF). But, instead of having tag to anchor distance as the input, I want to directly make the coordinate-xy as the input. Which parameter should be edited so that I can achieve the requirements below?

Input: XY-coordinate from UWB
Output: Filtered XY-coordinate from UWB

Thank you.