TY - JOUR
T1 - DRFM-BASED REPEATER JAMMING COGNITION METHOD BASED ON RESNET WITH CHANNEL-ATTENTION MECHANISM
AU - Zhang, Zhengyan
AU - Han, Bowen
AU - Liu, Fengrui
AU - Qu, Xiaodong
AU - Yang, Xiaopeng
N1 - Publisher Copyright:
© The Institution of Engineering & Technology 2023.
PY - 2023
Y1 - 2023
N2 - With the development of electronic countermeasure (ECM) technology, various kinds of digital radio frequency memory (DRFM)-based repeater jamming have been proposed, resulting in great influence on radar detection, localization and tracking. It is hard to obtain sufficient jamming samples, which brings difficulties to the jamming cognition. To solve this issue, a DRFM-based repeater jamming cognition method based on ResNet with channel-attention mechanism is proposed. In the proposed method, short-time Fourier transform (STFT) is utilized to generate the time-frequency maps of DRFM-based repeater jamming. Then, the 2D-wavelet transform is used to decompose the time-frequency maps and extract the texture feature. Moreover, the ResNet blocks with channel-attention mechanism are utilized to enhance the non-linear fitting ability of the network under small sample learning situation. Several simulations are conducted to illustrate that the proposed method achieves a higher classification accuracy for DRFM-based repeater jamming.
AB - With the development of electronic countermeasure (ECM) technology, various kinds of digital radio frequency memory (DRFM)-based repeater jamming have been proposed, resulting in great influence on radar detection, localization and tracking. It is hard to obtain sufficient jamming samples, which brings difficulties to the jamming cognition. To solve this issue, a DRFM-based repeater jamming cognition method based on ResNet with channel-attention mechanism is proposed. In the proposed method, short-time Fourier transform (STFT) is utilized to generate the time-frequency maps of DRFM-based repeater jamming. Then, the 2D-wavelet transform is used to decompose the time-frequency maps and extract the texture feature. Moreover, the ResNet blocks with channel-attention mechanism are utilized to enhance the non-linear fitting ability of the network under small sample learning situation. Several simulations are conducted to illustrate that the proposed method achieves a higher classification accuracy for DRFM-based repeater jamming.
KW - CHANNEL-ATTENTION MECHANISM
KW - DEEP LEARNING
KW - ELETRONIC COUNTER-COUNTERMENSURE
KW - JAMMING COGNITION
KW - TIME-FREQUENCY ANALYSIS
UR - http://www.scopus.com/inward/record.url?scp=85203145512&partnerID=8YFLogxK
U2 - 10.1049/icp.2024.1693
DO - 10.1049/icp.2024.1693
M3 - Conference article
AN - SCOPUS:85203145512
SN - 2732-4494
VL - 2023
SP - 3653
EP - 3658
JO - IET Conference Proceedings
JF - IET Conference Proceedings
IS - 47
T2 - IET International Radar Conference 2023, IRC 2023
Y2 - 3 December 2023 through 5 December 2023
ER -