为了提高认知无线电系统中低信噪比条件下的频谱感知性能,提出了基于非负矩阵分解的频谱感知方法。在无需知道被感知信号的先验信息的条件下,将原始信号进行短时傅里叶变换后,利用非负矩阵分解的噪声与信号之间的特征矩阵存在的差异性,将特征矩阵作为检测统计量进行频谱感知。仿真结果表明,基于非负矩阵分解的频谱感知方法在低信噪比条件下,具有较传统的能量检测方法与循环平稳检测方法更优的感知性能。
To improve the performance of spectrum sensing of cognitive radio system in low SNR,a spectrum sensing based on nonnegative matrix factorization is proposed. After transforming the original signal by STFT,by using the differences between signal and noise of basis matrices,the basis matrices can be used as test statistics to perform the spectrum sensing,without knowing the prior information of sensed signal. The simulation results show the proposed method has better performance compared with the traditional methods of Energy Detection and Cyclostationary Feature Detection in low SNR.