Journal of Advanced Navigation Technology (한국항행학회논문지)

The Korean Navigation Institute (한국항행학회)
권호 표지
DOI QR코드

DOI QR Code

Design of Radar Signal Processing System for Drone Detection

드론 검출을 위한 레이다 신호처리 시스템 설계

  • Hong-suk Kim (School of Electronics and Information Engineering, Korea Aerospace University) ;
  • Gyu-ri Ban (School of Electronics and Information Engineering, Korea Aerospace University) ;
  • Ji-hun Seo (School of Electronics and Information Engineering, Korea Aerospace University) ;
  • Yunho Jung (School of Electronics and Information Engineering, Korea Aerospace University)
  • 김홍석 (한국항공대학교 항공전자정보공학부) ;
  • 반규리 (한국항공대학교 항공전자정보공학부) ;
  • 서지훈 (한국항공대학교 항공전자정보공학부) ;
  • 정윤호 (한국항공대학교 항공전자정보공학부)
  • Received : 2024.09.24
  • Accepted : 2024.10.29
  • Published : 2024.10.31
Download PDF
⟨ Previous Next ⟩

Abstract

In this paper, we present the design and implementation results of a system that classifies drones from other objects using an FMCW (frequency-modulated continuous wave) radar sensor. The proposed system detects various objects through a four-stage signal processing procedure, consisting of FFT, CFAR, clustering, and tracking, using signals received from the radar sensor. Subsequently, a deep learning process is conducted to classify the detected objects as either drones or other objects. To mitigate the high computational demands and extensive memory requirements of deep learning, a BNN (binary neural network) structure was applied, binarizing the CNN (convolutional neural network) operations. The performance evaluation and verification results demonstrated a drone classification accuracy of 89.33%, with a total execution time of 4 ms, confirming the feasibility of real-time operation.

본 논문에서는 FMCW (frequency modulated continuous wave) 레이다 센서를 활용하여 드론 검출이 가능한 시스템 설계 및 구현 결과를 제시한다. 드론 검출 시스템의 구현을 위해, 레이다 센서로부터 입력된 신호를 FFT, CFAR, clustering, tracking으로 이어지는 총 4단계의 신호처리 과정을 통해 객체를 탐지하고, 해당 객체를 드론과 다른 사물로 분류하기 위한 딥러닝 추론 과정을 수행한다. 딥러닝의 높은 연산량과 많은 메모리 요구를 감소시키기 위해 CNN (convolution neural network) 연산을 이진화하여 수행하는 BNN (binary neural network) 구조를 적용하였다. 성능 평가 및 검증 결과 89.33%의 객체 구분 정확도를 확인할 수 있었고, 총 수행 시간은 4 ms로 실시간 동작이 가능함을 확인하였다.

Keywords

Acknowledgement

본 연구는 2024년도 정부 (과학기술정보통신부)의 재원으로 정보통신기획평가원의 지원을 받아 수행되었으며 (No. 2017-0-00528), CAD tool은 IDEC에 의해 지원되었음.

References

  1. C. Briese and L. Guenther, "Deep learning with semi-synthetic training images for detection of non-cooperative UAVs," in 2019 International Conference on Unmanned Aircraft Systems, Atlanta: GA, pp. 981-988, Jun. 2019. DOI: 10.1109/ICUAS.2019.8797731.
  2. M. Jung and S. Lee, "Object perception algorithm based on LiDAR for autonomous vehicle," in 2017 IEIE Fall Conference, Incheon: Korea, pp. 639-642, Nov. 2017. Retrieved from https://www.dbpia.co.kr/journal/articleDetail?nodeId=NODE07276399.
  3. C. J. Li and H. Ling, "An investigation on the radar signatures of small consumer drones," in IEEE Antennas and Wireless Propagation Letters, Vol. 16, pp. 649-652, Jul. 2017. DOI: 10.1109/LAWP.2016.2594766.
  4. X. Wang, H. Liang and P. Wang, "Detection and tracking of UAVs using interferometric radar," in 2019 International Radar Conference, Toulon: France, pp. 1-6, Sep. 2019. DOI: 10.1109/RADAR41533.2019.171312.
  5. H. Rohling and M. Kronauge, "New radar waveform based on a chirp sequence," in 2014 International Radar Conference, Lille: France, pp. 1-4, Oct. 2014. DOI: 10.1109/RADAR.2014.7060246.
  6. E. Hyun and J.-H. Lee, "Multi-target tracking scheme using a track management table for automotive radar systems," in 2016 17th International Radar Symposium, Krakow: Poland, pp.1-5, May. 2016. DOI: 10.1109/IRS.2016.7497283.
  7. L. Wang, J. Tang and Q. Liao, "A study on radar target detection based on deep neural networks," in IEEE Sensors Letters, Vol. 3, No. 3, pp. 1-4, Mar. 2019. DOI: 10.1109/LSENS.2019.2896072.
  8. J. H. Lee, Y. Yim, H. Choi, and Y. H. Jung, "Deep learning-based drone detection through sensor fusion technique of image sensor and FMCW radar," in 2020 IEIE Summer Conference, Jeju: Korea, pp. 690-691, Aug. 2020. Retrieved from https://www.dbpia.co.kr/journal/articleDetail?nodeId=NODE10447956.
  9. H. V. Quang, L. Kim, C. S. Hong, "A deep learning accelerator for a typical binary neural network layer," in Korea Computer Congress 2021, Jeju: Korea, pp.597-599, June 2021. Retrieved from https://www.dbpia.co.kr/Journal/articleDetail?nodeId=NODE10583012.
  10. B. R. Mahafza, Radar Systems Analysis and Design Using Matlab, 2nd ed. New York, NY: CRC Press, 2005.
  11. Y. -J. Kong, S. -K. Woo, S. Park, S. -Y. Shin, Y. H. Jang, E. Yang, "OS CFAR computation time reduction technique to apply radar system in real time," The Journal of Korean Institute of Electromagnetic Engineering and Science, Vol. 29, No. 10, pp. 791-798, Oct. 2018. DOI: https://doi.org/10.5515/KJKIEES.2018.29.10.791.
  12. M. Ester, H.-P. Kriegel, J. Sander, and X. Xu, "A density-based algorithm for discovering clusters in large spatial databases with noise," in Proceedings of the 2nd International Conference on Knowledge Discovery and Data Mining, Portland: Oregon, pp. 226-231, Aug. 1996. Retrieved from https://dl.acm.org/doi/10.5555/3001460.3001507.
  13. M. Courbariaux, I. Hubara, D. Soudry, R. El-Yaniv, and Y. Bengio, "Binarized neural networks: training deep neural networks with weights and activations constrained to +1 or -1," arXiv, Feb. 2016. DOI: https://doi.org/10.48550/arXiv.1602.02830.
  14. Raspberry Pi. raspberry-pi-4-model-b [Internet]. Available: https://www.raspberrypi.com/products/raspberry-pi-4-model-b/.
  15. Analog Devices. EVAL-DEMORAD [Internet]. Available: https://www.analog.com/en/resources/evaluation-hardware-and-software/evaluation-boards-kits/eval-demorad.html#eb-overview.
  16. Terasic Technologies. DE10-Lite Board [Internet]. Availabl e: https://www.terasic.com.tw/cgi-bin/page/archive.pl?Language=English&CategoryNo=234&No=1021.