LPI Radar Signals Modulation Recognition Based on ACDCA-ResNeXt.
Xudong Wang,Guiguang Xu,He Yan,Daiyin Zhu,Ying Wen,Zehu Luo
DOI: https://doi.org/10.1109/access.2023.3270231
IF: 3.9
2023-01-01
IEEE Access
Abstract:For low probability of intercept (LPI) radar waveform identification accuracy (ACC) problem at low Signal-to-Noise Ratios (SNRs), an approach based on time-frequency analysis (TFA) and Asymmetric Dilated Convolution Coordinate Attention Residual networks (ACDCA-ResNeXt) is proposed to recognize twelve kinds of LPI radar signals automatically. First, we apply Choi-Williams distribution (CWD), which shows superior performance at low SNRs, to transforming radar signals into time-frequency images (TFI). Then, in order to obtain the high-quality TFIs, a series of image processing techniques, including 2D Wiener filtering, image cutting, and image resize, are used to remove the background noise and redundant frequency bands of the TFI and obtain a fixed-size gray scale image containing main morphological features of the TFI. Finally, the TFIs are input into ACDCA-ResNeXt network that can extract and learn deep features to recognize radar waveforms. Furthermore, a fusion loss function, which is composed of a soft-label smoothed cross entropy loss function and a center loss function, improves the generalization capability performance of network and achieves a better clustering effect. Experimental results demonstrate that, for twelve kinds of LPI radar waveforms, the overall recognition ACC of the proposed approach achieves 97.94% when SNR is −8 dB.