Affiliation:
1. Shanghai Dianji University
2. Kanagawa University
Abstract
Abstract
In the face of the increasingly complex wireless communication environment, the traditional communication radiation source individual subtle feature recognition algorithm has the defects of poor real-time performance and low accuracy when dealing with the radiation source signal of small samples. Aiming at this problem, a subtle feature recognition method for communication radiation sources based on point density "painting" is proposed. Firstly, a signal feature extraction algorithm based on improved heat map is proposed to improve the accuracy and real-time performance of the traditional algorithm in extracting features when dealing with small sample signals. Second, based on the I/Q scatterplot, a fusion feature extraction algorithm based on the signal heatmap and scatterplot is proposed, which combines the region division strategy and the clustering fusion algorithm to extract the fusion features of the signal in an early fusion manner. Again, to address the problem of low recognition rate of fusion features, we introduce the concept of user portrait in the marketing field, and innovatively propose an algorithm for recognizing individual subtle features of communication radiation sources based on point density "portrait". Finally, a convolutional neural network (CNN) model is constructed to train the extracted point density image data of different classes of complex baseband signals, and the measured WIFI data collected in the darkroom environment of the laboratory is used to verify the effectiveness of the algorithm. The simulation results show that the algorithm can improve the accuracy by 1–4%, with a recognition rate of 100%, and exhibits good stability compared to the feature extraction algorithm based on the fusion feature map of signal heat map, I/Q scatter plot and signal heatmap .This provides an effective and reliable solution for the accurate identification of individual subtle features of communication radiation sources with small samples in complex environments.
Publisher
Research Square Platform LLC
Reference22 articles.
1. Chen, Y., Yu, L., Yao, Y., & Zhu, L. Individual Identification Technology of Communication Radiation Sources Based on Deep Learning, 2020 IEEE 20th International Conference on Communication Technology (ICCT), Nanning, China, 2020, pp. 1301–1305, 10.1109/ICCT50939.2020.9295728.
2. Guanghua Yi, X., Hao, X., Dai, Y. J., Liu, Y., & Han, Y. Automatic modulation recognition of radiation source signals based on two-dimensional data matrix and improved residual neural network, Defence Technology, 2023,ISSN 2214–9147, https://doi.org/10.1016/j.dt.2023.07.004.
3. Modulation recognition algorithm based on transfer meta-learning [J];Yiqiong PANG;Acta Armamentarii,2022
4. Distorted radar electromagnetic signal recognition based on meta-learning [J];Kang YAN;Journal of Electronics & Information Technology,2022
5. An emitter individual identification method for small samples based on optimized siamese networks [J];LIANG Xianming;Telecommunication Engineering,2022