한국컴퓨터정보학회논문지 (Journal of the Korea Society of Computer and Information)

한국컴퓨터정보학회 (Korean Society of Computer Information)
권호 표지
DOI QR코드

DOI QR Code

Radio Frequency-based Drone Detection and Classification Using Discrete Fourier Transform and LightGBM

  • Ki-Hyeon Sung (Dept. of Defense Science, Korea National Defense University) ;
  • Soo-Jin Lee (Dept. of Defense Science, Korea National Defense University)
  • 투고 : 2024.07.26
  • 심사 : 2024.09.25
  • 발행 : 2024.10.31
PDF 다운로드
⟨ 이전 논문 다음 논문 ⟩

초록

본 연구에서는 드론 및 관련 장치들로부터 생성되는 무선주파수 신호를 기반으로 드론을 탐지하고 기종을 식별하는 모델을 제안하였다. 전장 환경에서의 활용 가능성을 높이기 위해 모델은 경량화와 신속한 탐지를 최우선으로 고려하되 높은 탐지 정확도도 보장할 수 있도록 설계하였고, 데이터 전처리는 Hilbert-Huang 변환보다 처리 속도가 더 빠른 이산 푸리에 변환으로 수행하였다. 학습 모델은 비전문가도 쉽게 사용할 수 있고 분류 속도 및 정확도 측면에서도 탁월한 성능을 발휘하는 LightGBM 모델을 채택하였다. 제안하는 모델의 성능 검증은 공개 드론 무선주파수 데이터세트인 CardRF dataset을 활용하여 수행하였다. 실험결과 드론, WiFi 및 Bluetooth 장치 3종을 탐지 및 식별하는 3 클래스 다중분류의 정확도는 이산 푸리에 변환으로 전처리를 수행하는 과정에서 샘플 포인트 수를 100k로 설정한 경우 99.63%, 500k로 설정한 경우 99.40%로 나타났다. 드론 6종, Bluetooth 장치 2종 및 WiFi 장치 2종에 대한 10 클래스 다중분류 실험에서는 샘플 포인트 수가 100k인 경우 95.65%, 500k인 경우 96.83%를 달성하여 이전 연구 대비 상당히 향상된 탐지성능을 보임을 확인하였다.

In this study, we proposed an efficient model that can detect and classify the drones and related devices based on radio frequency signals. In order to increase the applicability in the battlefield, proposed model was designed to be lightweight, to ensure rapid detection and high detection accuracy. Data preprocessing was performed by applying a Discrete Fourier Transform (DFT) that is faster than Hilbert-Huang Transform (HHT). We adopted the LightGBM model as the learning model, which can be easily used by non-professionals and guarantees excellent performance in terms of classification speed and accuracy. CardRF dataset was used to verify the performance of the proposed model. As a result of the experiment, the accuracy of 3 classes classification for detecting and classifying drones, WiFi, and Bluetooth device was 99.63% when the number of sample points was set to 100k and 99.40% when set to 500k during the data preprocessing with DFT. And, in the 10 classes classification for 6 drones, 2 Bluetooth devices, and 2 WiFi devices, the accuracy was 95.65% for 100k and 96.83% for 500k, confirming significantly improved detection performance compared to previous studies.

키워드

참고문헌

  1. The Business Research Company, "UAV Drones Global Market Report 2024", https://www.researchandmarkets.com/report/unmanned-aerial-vehicles-uavs-drones
  2. K. Telli, O. Kraa, Y. Himeur, A. Ouamane, and M. Boumehraz, "A Comprehensive Review of Recent Research Trends on Unmanned Aerial Vehicles (UAVs)", Systems, 11, 400, Nov 2023. DOI:10.3390/systems11080400
  3. Anna Jackman, "Police Drones: Uses, Challenges, Futures", https://research.reading.ac.uk/drone-geographies/wp-content/uploads/sites/271/2023/09/Police-report.pdf#:~:text=URL%3A%20https%3A%2F%2Fresearch.reading.ac.uk%2Fdrone
  4. O. H. Dahle, J. Rydberg, M. Dullweber, N. Peinecke and A. A. A. Bechina, "A proposal for a common metric for drone traffic density", International Conference on Unmanned Aircraft Systems (ICUAS), pp. 64-72, Dubrovnik, Croatia, Jun 2022. DOI: 10.1109/ICUAS54217.2022.9836143.
  5. Seth Cropsey, "Drone Warfare in Ukraine: Historical Context and Implications for the Future," https://www.hoover.org/research/drone-warfare-ukraine-historical-context-and-implications-future
  6. Lauren Kahn, "How Ukraine Is Using Drones Against Russia," https://www.cfr.org/in-brief/how-ukraine-using-drones-against-russia
  7. Kristen D. Thompson, "How the Drone War in Ukraine Is Transforming Conflict," https://www.cfr.org/article/how-dronewar-ukraine-transforming-conflict
  8. M. H. Rahman, Mohammad A. S. Sejan, M. A. Aziz, R. Tabassum, Jung-In Baik, and Hyoung-Kyu Song, "A Comprehensive Survey of Unmanned Aerial Vehicles Detection and Classification Using Machine Learning Approach: Challenges, Solutions, and Future Directions", 2024 Remote Sensing 16, no. 5: 879., Mar 2024. DOI:10.3390/rs16050879
  9. Olusiji O. Medaiyese, M. Ezuma, A. Lauf, and A. Adeniran, "Cardinal RF (CardRF): An Outdoor UAV/UAS/Drone RF Signals with Bluetooth and WiFi Signals Dataset", IEEE Dataport, Jul 2022. DOI:10.21227/1xp7-ge95
  10. U. Seidaliyeva, L. Ilipbayeva, K. Taissariyeva, N. Smailov, and E. Matson, "Advances and Challenges in Drone Detection and Classification Techniques: A State-of-the-Art Review", Sensors 24(1):125, Dec 2023. DOI: 10.3390/ s24010125.
  11. N. Al-lQubaydhi, A. Alenezi, T. Alanazi, A. Senyor, N. Alanezi, B. Alotaibi, M. Alotaibi, A. Razaque, and S. Hariri, "Deep learning for unmanned aerial vehicles detection: A review", Computer Science Review, Vol. 51, No. C, Jun 2024. DOI: 10.1016/j.cosrev.2023.100614.
  12. Olusiji O. Medaiyese, M. Ezuma, A. P. Lauf, and A. A. Adeniran, "Hierarchical Learning Framework for UAV Detection and Identification", IEEE Journal of Radio Frequency Identification, vol. 6, pp. 176-188, Mar 2022. DOI: 10.1109/JRFID.2022.3157653.
  13. Olusiji O. Medaiyese, M. Ezuma, A. P. Lauf, and Ismail Guvenc, "Wavelet transform analytics for RF-based UAV detection and identification system using machine learning", Pervasive and Mobile Computing, Volume 82, 101569, Jun 2022. DOI:10.1016/j.pmcj.2022.101569.
  14. Alam, Syed Samiul, Arbil Chakma, Md Habibur Rahman, Raihan Bin Mofidul, Md Morshed Alam, Ida Bagus Krishna Yoga Utama, and Yeong Min Jang, "RF-Enabled Deep-Learning-Assisted Drone Detection and Identification: An End-to-End Approach", Sensors 23, no. 9: 4202, April 2023. DOI:10.3390/s23094202
  15. Yan, Xiang, Bing Han, Zhigang Su, Jingtang Hao., "SignalFormer: Hybrid Transformer for Automatic Drone Identification Based on Drone RF Signals", Sensors 23, no. 22: 9098, November 2023. DOI:10.3390/s23229098
  16. Mohammad F. Al-Sa'd, Abdulla Al-Ali, Amr Mohamed, Tamer Khattab, and Aiman Erbad, "RF-based drone detection and identification using deep learning approaches: An initiative towards a large open source drone database", Future Generation Computer Systems, 100: 86-97, November, 2019. DOI:10.1016/j.future.2019.05.007
  17. M. S. Allahham, T. Khattab, and A. Mohamed, "Deep Learning for RF-Based Drone Detection and Identification: A Multi-Channel 1-D Convolutional Neural Networks Approach", 2020 IEEE International Conference on Informatics, IoT, and Enabling Technologies (ICIoT), pp. 112-117, Doha, Qatar, February, 2020. DOI:10.1109/ICIoT48696.2020.9089657.
  18. Z. Oubrahim, A. Yassine, B. Mohamed, and O. Mohammed, "Power Quality Disturbances Characterization Using Signal Processing and Pattern Recognition Techniques: A Comprehensive Review", Energies 16, no. 6: 2685, Mar 2023. DOI:10.3390/en16062685
  19. O. Brandes, J. Farley, M. Hinich, and U. Zackrisson, "The time domain and the frequency domain in time series analysis", The Swedish Journal of Economics, 25-42, Mar 1968. DOI:10.2307/3438983
  20. I. Nemer, S. Tarek, A. Irfan, Ansar Ul-Haque Yasar, and Mohammad A. R. Abdeen., "RF-Based UAV Detection and Identification Using Hierarchical Learning Approach", Sensors 21, no. 6: 1947, Mar 2021. DOI:10.3390/s21061947
  21. Norden E. Huang, Zheng Shen, Steven R. Long, Manli C. Wu, Hsing H. Shih, Quanan Zheng, Nai-Chyuan Yen, Chi Chao Tung and Henry H. Liu, "The empirical mode decomposition and the Hilbert spectrum for nonlinear and non-stationary time series analysis", Proc. R. Soc. Lond. A.454903-995 March 1998. DOI:10.1098/rspa.1998.0193
  22. Guolin Ke, Qi Meng, Thomas Finley, Taifeng Wang, Wei Chen, Weidong Ma, Qiwei Ye, and Tie-Yan Liu, "LightGBM: A highly efficient gradient boosting decision tree", Advances in neural information processing systems 30, 2017.
  23. C. Candan, "Analysis and Further Improvement of Fine Resolution Frequency Estimation Method From Three DFT Samples", IEEE Signal Processing Letters, Vol. 20, No. 9, pp. 913-916, Sep 2013. DOI: 10.1109/LSP.2013.2273616.