Academic Bibliography

 Search 200 years of publications by Ghent University researchers.
Advanced search
1 file | 2.10 MB
Add to list
  1. Search results
  2. Adapting UWB AoA estimation towards u...

Adapting UWB AoA estimation towards unseen environments using transfer learning and data augmentation

Mostafa Naseri (UGent) , Adnan Shahid (UGent) and Eli De Poorter (UGent)
(2024) INTERNET OF THINGS. 27.
Author
Organization
Project
Abstract
Ultra-wideband technology has become increasingly prevalent in localization systems, particularly with the emergence of multi-antenna systems capable of estimating both distance and angle of arrival (AoA) for incoming signals. However, most scientific research analyzes the accuracy of AoA in one specific environment for which the estimator is trained. In this paper, we analyze the performance of various AoA estimation algorithms, such as deep convolutional neural networks (DCNN), multiple signal classification (MUSIC), and phase difference of arrival (PDoA), in unseen environments. Prior work already demonstrated the superior performance of ML for AoA estimation compared to PDoA or MUSIC. We show that MUSIC, PDoA and ML solutions suffer from degradation in unseen environments at the 90th percentile of error, with ML-based AoA estimation degrading by about 14 degrees in unseen environments compared to 4 degrees for PDoA. We demonstrate that while PDoA more effectively corrects AoA at the median level in unseen environments, ML-based methods excel at correcting higher-percentile AoA errors, including outliers. Finally, we propose a novel framework to further improve the correction of AoA outliers for DCNN-based AoA estimators using data augmentation and transfer learning, resulting in a median angular error of only 5 degrees in unseen environments, even considering a field of view up to 90 degrees.
Keywords
Transfer Learning (TL), Machine learning (ML), Generalization, Ultra-wideband (UWB), Data augmentation (DA), NEURAL-NETWORKS, LOCALIZATION

Downloads

  • Download
    (...).pdf
    • full text (Published version)
    • |
    • UGent only
    • |
    • PDF
    • |
    • 2.10 MB

Citation

Please use this url to cite or link to this publication:

MLA
Naseri, Mostafa, et al. “Adapting UWB AoA Estimation towards Unseen Environments Using Transfer Learning and Data Augmentation.” INTERNET OF THINGS, vol. 27, 2024, doi:10.1016/j.iot.2024.101298.
APA
Naseri, M., Shahid, A., & De Poorter, E. (2024). Adapting UWB AoA estimation towards unseen environments using transfer learning and data augmentation. INTERNET OF THINGS, 27. https://doi.org/10.1016/j.iot.2024.101298
Chicago author-date
Naseri, Mostafa, Adnan Shahid, and Eli De Poorter. 2024. “Adapting UWB AoA Estimation towards Unseen Environments Using Transfer Learning and Data Augmentation.” INTERNET OF THINGS 27. https://doi.org/10.1016/j.iot.2024.101298.
Chicago author-date (all authors)
Naseri, Mostafa, Adnan Shahid, and Eli De Poorter. 2024. “Adapting UWB AoA Estimation towards Unseen Environments Using Transfer Learning and Data Augmentation.” INTERNET OF THINGS 27. doi:10.1016/j.iot.2024.101298.
Vancouver
1.
Naseri M, Shahid A, De Poorter E. Adapting UWB AoA estimation towards unseen environments using transfer learning and data augmentation. INTERNET OF THINGS. 2024;27.
IEEE
[1]
M. Naseri, A. Shahid, and E. De Poorter, “Adapting UWB AoA estimation towards unseen environments using transfer learning and data augmentation,” INTERNET OF THINGS, vol. 27, 2024.
@article{01J4R9PH5Q14SJBM2AYJNRKGWX,
  abstract     = {{Ultra-wideband technology has become increasingly prevalent in localization systems, particularly with the emergence of multi-antenna systems capable of estimating both distance and angle of arrival (AoA) for incoming signals. However, most scientific research analyzes the accuracy of AoA in one specific environment for which the estimator is trained. In this paper, we analyze the performance of various AoA estimation algorithms, such as deep convolutional neural networks (DCNN), multiple signal classification (MUSIC), and phase difference of arrival (PDoA), in unseen environments. Prior work already demonstrated the superior performance of ML for AoA estimation compared to PDoA or MUSIC. We show that MUSIC, PDoA and ML solutions suffer from degradation in unseen environments at the 90th percentile of error, with ML-based AoA estimation degrading by about 14 degrees in unseen environments compared to 4 degrees for PDoA. We demonstrate that while PDoA more effectively corrects AoA at the median level in unseen environments, ML-based methods excel at correcting higher-percentile AoA errors, including outliers. Finally, we propose a novel framework to further improve the correction of AoA outliers for DCNN-based AoA estimators using data augmentation and transfer learning, resulting in a median angular error of only 5 degrees in unseen environments, even considering a field of view up to 90 degrees.}},
  articleno    = {{101298}},
  author       = {{Naseri, Mostafa and Shahid, Adnan and De Poorter, Eli}},
  issn         = {{2543-1536}},
  journal      = {{INTERNET OF THINGS}},
  keywords     = {{Transfer Learning (TL),Machine learning (ML),Generalization,Ultra-wideband (UWB),Data augmentation (DA),NEURAL-NETWORKS,LOCALIZATION}},
  language     = {{eng}},
  pages        = {{15}},
  title        = {{Adapting UWB AoA estimation towards unseen environments using transfer learning and data augmentation}},
  url          = {{http://doi.org/10.1016/j.iot.2024.101298}},
  volume       = {{27}},
  year         = {{2024}},
}
Altmetric
View in Altmetric
Web of Science
Times cited: