基于稀疏傅里叶变换的快速频谱感知方法  

Fast spectrum sensing method based on sparse FFT

作　　者：陈李 叶芃[1] 杨慧芝 杨扩军[1] CHEN Li;YE Peng;YANG Huizhi;YANG Kuojun(School of Automation Engineering,University of Electronic Science and Technology of China,Chengdu 611731,China;Sciences Basic Teaching Department,Sichuan College of Architectural Technology,Chengdu 610300,China)

机构地区：[1]电子科技大学自动化工程学院,成都611731 [2]四川建筑职业技术学院基础教学部,成都610300

出　　处：《电子科技大学学报》2025年第2期274-279,共6页Journal of University of Electronic Science and Technology of China

基　　金：国家自然科学基金(62371097)。

摘　　要：在通信、雷达等应用场景中,常常需要对稀疏信号进行高精度的频谱运算。传统基于快速傅里叶变换的方法需要大量运算资源,频谱分析效率低。为了解决高精度和实时性的矛盾,该文提出了一种基于稀疏傅里叶变换的频谱分析方法,利用延时采样的相位旋转效应,在低采样率下实现宽带信号的快速频谱感知。实验结果显示,这种方法在欠采样且存在混叠的稀疏信号测试场景下大幅减少了运算压力,运算效率比FFT提升了2倍以上,在典型的稀疏场景下,对信号恢复精度超过95%。In communication,radar,and other applications,it is often necessary to perform high-precisionspectral computations on sparse signals.Traditional methods based on Fast Fourier Transform(FFT)requiresubstantial computational resources,leading to a decrease in the efficiency of spectral analysis.To resolve theconflict between high precision and real-time requirements,this paper proposes a spectral analysis method based onSparse Fourier Transform(SFT).By utilizing the phase rotation effect of delayed sampling,this method achievesrapid spectral perception of wideband signals at low sampling rates.Experimental results show that this approachsignificantly reduces computational burden in under-sampled and aliased sparse signal testing scenarios,improvingcomputational efficiency by more than 2 times compared to FFT.In typical sparse scenarios,the signal recoveryaccuracy exceeds 95%.

关 键 词：稀疏信号 傅里叶变换 频谱感知 可重构采样 高速数据采集系统 

分 类 号：TH85[机械工程—仪器科学与技术]