定频干扰下基于奇异值熵的跳频信号检测  

Frequency hopping signal detection using singular value entropy under fixed frequency interference

作　　者：齐雨 张晓林 孙溶辰 QI Yu;ZHANG Xiaolin;SUN Rongchen(College of Information and Communication Engineering,Harbin Engineering University,Harbin 150001,China)

机构地区：[1]哈尔滨工程大学信息与通信工程学院,黑龙江哈尔滨150001

出　　处：《哈尔滨工程大学学报》2025年第2期355-362,共8页Journal of Harbin Engineering University

基　　金：国家自然科学基金项目(62001139)。

摘　　要：针对现有跳频信号检测算法在低信噪比下性能急剧下降的问题,本文利用定频信号、跳频信号和噪声时频图像的特征差异,提出基于奇异值熵的跳频信号盲检测算法。该算法经过时频对消方法剔除时频图中的定频干扰分量,通过时频迁移及奇异值分解计算奇异值熵反映时频图的能量分布特征,采用奇异值熵作为跳频信号的检测量,实现了定频干扰下跳频信号的准确检测。仿真结果表明:所提算法能在信噪比高于-14 dB时保持100%的检测概率,在定频干扰的频率与信号重叠时也具有良好的检测性能,还可用于高斯白噪声背景下定频信号和跳频信号的识别。To address the considerable drop in performance of existing frequency hopping signal detection algorithms at low signal-to-noise ratios,this paper presents a blind detection algorithm using singular value entropy.This algorithm removes fixed-frequency interference components from the time-frequency image using the time-frequency cancellation method.Then,it employs time-frequency migration and singular value decomposition to compute singular value entropy,which reflects the energy distribution characteristics of the time-frequency image.Singular value entropy serves as the detection statistic for frequency-hopping signals,facilitating accurate detection under fixed frequency interference.Simulation results demonstrate that the proposed algorithm achieves 100%detection probability when the signal-to-noise ratio exceeds-14 dB,exhibiting strong performance even when the frequency of the fixed-frequency interference overlaps with the signal.It can also be used for identifying fixed-frequency signals and frequency-hopping signals against a background of Gaussian white noise.

关 键 词：跳频信号 短时傅里叶变换 时频对消 时频迁移 奇异值分解 奇异值熵 盲检测 定频干扰 高斯白噪声 

分 类 号：TN971[电子电信—信号与信息处理]