卫星频谱信号智能挖掘技术  

Intelligent Excavation Technologies for Satellite Spectrum Signals

作　　者：吕秋霖 丁晓进 张更新[2,3] LYU Qiulin;DING Xiaojin;ZHANG Gengxin(College of Internet of Things,Nanjing University of Post and Telecommunications,Nanjing 210003,China;College of Telecommunications&Information Engineering,Nanjing University of Post and Telecommunications,Nanjing 210003,China;Satellite Communication Research Institute,Nanjing University of Post and Telecommunications,Nanjing 210003,China)

机构地区：[1]南京邮电大学物联网学院,江苏南京210003 [2]南京邮电大学通信与信息工程学院,江苏南京210003 [3]南京邮电大学卫星通信研究所,江苏南京210003

出　　处：《天地一体化信息网络》2022年第3期30-36,共7页Space-Integrated-Ground Information Networks

基　　金：国家自然科学基金资助项目(No.62171234,No.U21A20450,No.61971440);江苏省前沿引领技术基础研究专项(No.BK20192002)。

摘　　要：随着空间信息网络的建设,可用频谱资源日益紧缺,面临的电磁环境越来越复杂,实现对全球电磁频谱态势的安全掌控所面临的挑战更加突出。针对以上问题,研究电磁频谱占用状态感知方法,通过构建时间卷积神经网络来感知频谱占用状态;探索频谱参数提取方法,结合底噪拟合和聚类分析,实现对中心频率、带宽、峰值功率等参数的提取;为挖掘异常频谱数据,研究利用神经网络自动检测起伏底噪、频谱草和大带宽等异常频谱;考虑星地间的大传播时延,进一步通过构建神经网络来实现对频谱占用状态的预测;将卫星频谱信号的感知、异常、参数和预测结果进行可视化呈现。初步评估结果表明,所设计的一套频谱感知、异常检测、参数认知和态势预测等卫星频谱智能挖掘技术,在提升频谱态势安全掌控能力的同时,还能有效提升频谱资源利用率。With the construction of spatial information network, the available spectrum resources are increasingly scarce, and the electromagnetic environment is becoming more and more complex, resulting a more prominent challenge of realizing the safe control of the global electromagnetic. Therefore, the sensing method of electromagnetic spectrum occupation state was studied, a time convolution neural network to sense the spectrum occupation state was constructed. To mine abnormal spectrum data, neural network was used to automatically detect abnormal spectrum, such as fl uctuation noise fl oor, spectrum grass and large bandwidth. The spectrum parameter extraction method was explored. Considering the large propagation delay between the spectrum-sensing satellite and its gateway, a neural network was constructed to predicted the spectrum occupation state. The perception, anomaly, parameters and prediction results of satellite spectrum signal were visualized. The preliminary evaluation results showed that the designed satellite-spectrum intelligent mining technologies could not only improved the sensing ability of spectrum situation, but also ef fectively improved the utilization of spectrum resources.

关 键 词：频谱感知 参数认知 异常检测 频谱态势 

分 类 号：V19[航空宇航科学与技术—人机与环境工程]