Communication Interference Intelligent Recognition in the Air-to-ground Collaboration Scenario
-
摘要: 针对现有通信干扰智能识别方法在小样本条件下识别精度低、网络模型欠拟合的问题，并形成通信干扰识别的空中与地面布设能力，该文提出一种空地协同场景下基于孪生网络的通信干扰智能识别方法。首先在空中无人机与地面设备之间构建空地协同的通信干扰认知架构，并通过提取所接收的通信干扰信号的时频图、分数阶傅里叶变换和星座图，对通信干扰信号进行智能表征，以作为网络的输入。然后搭建基于密集连接网络的网络结构，并设计双输入权值共享的孪生网络。最后，利用随机样本对孪生网络进行训练，并通过孪生单边网络构建基准通信干扰类型特征库进而实现通信干扰的智能识别。该方法通过度量两个样本之间的特征距离来判断样本的相似性，并通过相似度度量扩大了训练样本数量并训练了孪生网络模型。仿真结果表明，所提方法不但在较小数据集的条件下可有效地实现通信干扰的智能识别，而且相比现有的智能识别方法，所提方法的识别性能显著提升。Abstract: In view of the problems that the existing communication interference intelligent recognition methods have low recognition accuracy under the small samples condition and the under-fitting of the network model, an intelligent communication interference recognition method based on twin network in the air-to-ground collaboration scenario is proposed to form the air and ground layout ability. Firstly, the time-frequency diagram, fractional Fourier transform and constellation diagram of the received communication interference signals are extracted as network inputs in the air-to-ground collaboration communication interference cognitive architecture between unmanned air vehicles and ground equipment. Secondly, the network structure based on densely connected convolutional networks is built, and the twin network with dual input weight sharing is designed. Finally, the twin network is trained with random samples, and the benchmark communication interference type feature library is constructed through the twin unilateral network, so as to realize the communication interference intelligent identification. The proposed method evaluates the similarity of samples by measuring the feature distance between two samples, and expands the number of training samples and trains the Siamese network model through the similarity measurement. Simulation results show that the proposed method can effectively realize the communication interference recognition under the support of small data sets, and the recognition performance is significantly improved compared with the existing intelligent recognition methods.
-
[1] JOHNSTON J, LI Yinchuan, LOPS M, et al. ADMM-net for communication interference removal in stepped-frequency radar[J]. IEEE Transactions on Signal Processing, 2021, 69: 2818–2832. doi: 10.1109/TSP.2021.3076900 [2] KE Chenxi, LI Jingwen, CHENG Wei, et al. An intelligent anti-interference communication method based on game learning[C]. The 2021 IEEE 5th Advanced Information Technology, Electronic and Automation Control Conference, Chongqing, China, 2021: 182–186. [3] 黄国策, 王桂胜, 任清华, 等. 基于Hilbert信号空间的未知干扰自适应识别方法[J]. 电子与信息学报, 2019, 41(8): 1916–1923. doi: 10.11999/JEIT180891HUANG Guoce, WANG Guisheng, REN Qinghua, et al. Adaptive recognition method for unknown interference based on Hilbert signal space[J]. Journal of Electronics &Information Technology, 2019, 41(8): 1916–1923. doi: 10.11999/JEIT180891 [4] 冯熳, 王梓楠. 基于奇异值分解与神经网络的干扰识别[J]. 电子与信息学报, 2020, 42(11): 2573–2578. doi: 10.11999/JEIT190228FENG Man and WANG Zinan. Interference recognition based on singular value decomposition and neural network[J]. Journal of Electronics &Information Technology, 2020, 42(11): 2573–2578. doi: 10.11999/JEIT190228 [5] HUANG Liang, ZHANG You, PAN Weijian, et al. Visualizing deep learning-based radio modulation classifier[J]. IEEE Transactions on Cognitive Communications and Networking, 2021, 7(1): 47–58. doi: 10.1109/TCCN.2020.3048113 [6] 党泽. 基于深度学习的无线通信干扰信号识别与处理技术研究[D]. [硕士论文], 电子科技大学, 2020.DANG Ze. Research on the technology of wireless communication interference signal identification and processing based on deep learning[D]. [Master dissertation], University of Electronic Science and Technology of China, 2020. [7] GOODFELLOW I J, POUGET-ABADIE J, MIRZA M, et al. Generative adversarial nets[C]. The 27th International Conference on Neural Information Processing Systems, Montreal, Canada, 2014: 2672–2680. [8] KOCH G, ZEMEL R, and SALAKHUTDINOV R. Siamese neural networks for one-shot image recognition[C]. The 32nd International Conference on Machine Learning, Lille, France, 2015: 1–8. [9] CHOPRA S, HADSELL R, and LECUN Y. Learning a similarity metric discriminatively, with application to face verification[C]. 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, San Diego, USA, 2005: 539–546. [10] NOROUZI M, FLEET D J, and SALAKHUTDINOV R. Hamming distance metric learning[C]. The 25th International Conference on Neural Information Processing Systems, Lake Tahoe, USA, 2012: 1061–1069. [11] VINYALS O, BLUNDELL C, LILLICRAP T, et al. Matching networks for one shot learning[C]. Proceedings of the 30th International Conference on Neural Information Processing Systems, Barcelona, Spain, 2016: 3637–3645. doi: 10.5555/3157382.3157504. [12] XU Lu, YIN Xingyao, ZONG Zhaoyun, et al. Synchrosqueezing matching pursuit time–frequency analysis[J]. IEEE Geoscience and Remote Sensing Letters, 2021, 18(3): 411–415. doi: 10.1109/LGRS.2020.2978877 [13] SHI Jun, ZHENG Jiabin, LIU Xiaoping, et al. Novel short-time fractional Fourier transform: Theory, implementation, and applications[J]. IEEE Transactions on Signal Processing, 2020, 68: 3280–3295. doi: 10.1109/TSP.2020.2992865 [14] CHICCO D. Siamese Neural Networks: An Overview[M]. New York: Humana, 2021: 73–94. [15] YANG Ning, ZHANG Bangning, DING Guoru, et al. Specific emitter identification with limited samples: A model-agnostic meta-learning approach[J]. IEEE Communications Letters, 2022, 26(2): 345–349. doi: 10.1109/LCOMM.2021.3110775 -
下载:
图(10) / 表(1)
计量
- 文章访问数: 678
- HTML全文浏览量: 835
- PDF下载量: 187
- 被引次数: 0
