联合相位谱与幅度谱的时频图干扰检测和样式识别  

Interference Detection and Pattern Recognition Based on Time-Frequency Graph with Joint Phase and Amplitude Spectrum

作　　者：李扬清 陈章烜 洪江涛 宋沅隆 高秋雅 张志龙[2] LI Yangqing;CHEN Zhangxuan;HONG Jiangtao;SONG Yuanlong;GAO Qiuya;ZHANG Zhilong(The 7th Research Institute of China Electronics Technology Group Corporation,Guangzhou 510310,China;Beijing University of Posts and Telecommunications,Beijing 100876,China)

机构地区：[1]中国电子科技集团公司第七研究所,广东广州510310 [2]北京邮电大学,北京100876

出　　处：《移动通信》2023年第10期93-98,110,共7页Mobile Communications

摘　　要：在无线通信系统中,干扰信号检测和干扰样式识别作为电磁频谱监测系统的重要应用组成,在干扰认知和抗干扰方面起到非常关键的作用。为了提高检测和识别的准确率,提出一种基于联合相位谱图与幅度谱图的时频图深度目标检测识别方法。在传统的时频幅度谱图的基础上,提出采用差分相位谱的三角变换来构建无相位跳变和相位模糊的相位谱图,并采用基于图像的深度目标检测算法对联合相位谱图与幅度谱图的时频图像进行训练、识别与检测。仿真结果表明,结合幅度谱图与相位谱图的图像深度目标检测算法能够很好地检测干扰并识别干扰样式,相比于单纯的幅度谱检测方式在仿真合成数据集中的平均识别精度提升了5.9%。In wireless communication systems,interference signal detection and interference pattern recognition,as significant components of electromagnetic spectrum monitoring system,play a key role in interference cognition and anti-interference.In order to improve the accuracy of detection and recognition,a deep target detection and recognition method is proposed based on the time-frequency graph which combines both phase spectrogram and amplitude spectrogram.Based on traditional timefrequency amplitude spectrogram,a triangular transformation of the differential phase spectrum is proposed to construct a phase spectrogram without phase hop and ambiguity,and the image-based deep target detection algorithm is adopted to train,recognize and detect the time-frequency graph with the joint phase and amplitude spectrum.The simulation results show that our proposed deep target detection algorithm performs well in detecting the interference and recognizing the interference pattern.Compared with the simple amplitude spectrum detection method,the average recognition accuracy is improved by 5.9%in the simulation synthesis dataset.

关 键 词：干扰检测 干扰样式识别 时频图 相位谱图 图像目标检测 

分 类 号：TN929.5[电子电信—通信与信息系统]