| 摘要: |
| 针对当前低信噪比环境下，基于射频信号的无人机型号和飞行模式识别率低的问题，本文提出了一种基于MobileNet-DOA模型的无人机射频信号识别方法。该方法首先对原始无人机射频信号进行基于变分模态分解的信号预处理，降低背景噪声和同频干扰，然后利用短时傅里叶变换将预处理信号转换为时频图，最后利用MobileNet-DOA模型完成无人机射频信号识别。在模型方面，本文首先将DOConv卷积融合到MobileNetv4模型中，在增强模型特征提取能力的同时，提高了训练和运算速度。其次，使用FA注意力机制进一步提升了模型在低信噪比环境下的识别准确率。实验结果表明，该方法在-15dB到15dB信噪比范围内的平均检测准确率达到了94.83%，可应用于无人机实时检测识别系统中。 |
| 关键词: 无人机射频信号识别 变分模态分解 MobileNet模型 DOConv卷积 FA注意力机制 |
| DOI： |
| 分类号:TN911.7 |
| 基金项目:国家自然科学基金(62173124);国家自然科学基金(62101563)；河北省自然科学基金(F2022202064) |
|
| Drone Radio Frequency Signal Identification Based on MobileNet-DOA |
|
|
|
|
| Abstract: |
| To address the issue of low identification rates for drone models and flight modes based on radio frequency signals in low SNR environments, this paper proposed a drone radio frequency signal identification method based on the MobileNet-DOA model. The method first preprocessed the raw drone radio frequency signals using variational mode decomposition to reduce background noise and co-frequency interference. Then, the preprocessed signals were transformed into time-frequency images using short-time Fourier transform. Finally, the MobileNet-DOA model was employed to complete the identification of drone radio frequency signals. In terms of the model, this paper integrated DOConv convolution into the MobileNetv4 model, enhancing the feature extraction capability while improving training and computation speed. Additionally, the FA attention mechanism was utilized to further improve the identification accuracy of the model in low SNR environments. Experimental results showed that the proposed method achieved an average detection accuracy of 94.83% in the SNR range from -15 dB to 15 dB, making it applicable to real-time drone detection and identification systems. |
| Key words: drone radio frequency signal identification variational mode decomposition MobileNet DOConv FA attention mechanism |