EUVIMEDEUVIMED entdecken
Crossref

EKF-based multi-sensor fusion for UAV navigation: a comparative simulation study using GPS, IMU, and radar

D.C. Quang, P.V. Tran, D.S. Nguyen, D.L. Nguyen, H.N. Le, N. Kongprasert

Journal of Achievements in Materials and Manufacturing Engineering · 01.03.2026

DOI: 10.5604/01.3001.0055.7993 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Journal of Achievements in Materials and Manufacturing Engineering, erschienen am 01.03.2026. Verfasst von D.C. Quang und 5 weiteren Autor:innen.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Many uncrewed aerial vehicle (UAV) navigation systems rely heavily on Global Positioning System (GPS) and Inertial Measurement Unit (IMU) measurements, which may suffer from noise, drift, or temporary degradation. This study aims to improve UAV localisation performance by evaluating an Extended Kalman Filter (EKF)-based multi-sensor fusion framework that combines GPS, IMU, and radar measurements under a controlled simulation environment.



An EKF-based state-estimation framework is implemented to fuse measurements from multiple sensors to localise a UAV. Rather than proposing a new EKF formulation, the study compares five sensing configurations: GPS-only, radar-only, GPS+IMU, GPS+radar, and GPS+IMU+radar. Simulated UAV trajectory data are generated in MATLAB using a predefined figure-eight path, and localisation performance is assessed using the Root Mean Square Error (RMSE) metric.



The results show that all multi-sensor configurations improve localisation accuracy compared with the single-sensor baselines. Among the evaluated cases, the GPS+IMU+radar configuration achieves the best performance, with an average point-to-point error of 0.11 m, while the radar-only case yields the largest error of 0.84 m. The results indicate that integrating complementary sensors into the EKF framework can improve estimation accuracy and yield more consistent localisation performance under the simulated conditions examined in the given study.



The validation is limited to simulation-based experiments using a single representative trajectory and predefined sensor parameters. Therefore, the findings should be interpreted within the scope of controlled test conditions. Future work should examine more challenging scenarios, including measurement disturbances, sensor dropouts, bias effects, multiple trajectory patterns, and real-time or real-world flight validation.



The study provides a practical reference for selecting sensor combinations in UAV navigation systems by highlighting the trade-off between localisation accuracy, system complexity, and implementation cost. In particular, GPS+IMU+radar offers the highest accuracy in the present study, while GPS+radar provides a useful balance between performance and system complexity.



The contribution of this work lies in a systematic comparative evaluation of five sensor configurations within a unified EKF-based UAV localisation framework. The study offers application-oriented insight into the relative performance of GPS, IMU, and radar combinations and clarifies the practical benefits of tri-sensor fusion under a common simulation setting.

Zitieren

APA

Quang, D., Tran, P., Nguyen, D., Nguyen, D., Le, H., & Kongprasert, N. (2026). EKF-based multi-sensor fusion for UAV navigation: a comparative simulation study using GPS, IMU, and radar. Journal of Achievements in Materials and Manufacturing Engineering. https://doi.org/10.5604/01.3001.0055.7993

Vancouver

Quang D, Tran P, Nguyen D, Nguyen D, Le H, Kongprasert N. EKF-based multi-sensor fusion for UAV navigation: a comparative simulation study using GPS, IMU, and radar. Journal of Achievements in Materials and Manufacturing Engineering. 2026. doi:10.5604/01.3001.0055.7993

BibTeX
@article{Quang2026,
  author = {Quang, D.C. and Tran, P.V. and Nguyen, D.S. and Nguyen, D.L. and Le, H.N. and Kongprasert, N.},
  title = {EKF-based multi-sensor fusion for UAV navigation: a comparative simulation study using GPS, IMU, and radar},
  journal = {Journal of Achievements in Materials and Manufacturing Engineering},
  year = {2026},
  doi = {10.5604/01.3001.0055.7993}
}
Kostenlos starten →