基于混合时-频分布的信号检测与瞬时频率估计

Signal Detection and its Instantaneous Frequency Estimation Based on Hybrid Time-frequency Representations

提出了一种基于分数阶线性及分数阶非线性时-频分布混合法,该方法将分数阶维格拉分布与加窗分数阶短时傅里叶变换相结合,在适当的分数阶时-频域进行信号检测与瞬时频率的估计。对于多分量信号,该方法能够有效地消除交叉项,同时提高自项的时-频分辨率。在对线性调频信号检测与瞬时频率估计应用中,计算机仿真表明该方法是有效的,具有良好的工程应用前景。

A combined method based on fractional linear and fractional non-linear Time-Frequency Representations （TFRs） is proposed in this paper. This method combines the fractional Wigner-Ville Distribution （WVD） with the windowed fractional short-time Fourier transform to estimate the Instantaneous Frequency （IF） of signals in the appropriate fractional time-frequency domain. For a multi-component signal, the method can eliminate the cross terms, significantly and improve the time-frequency resolution of the auto-terms. It is applied to the detection and parameter estimation of Linear Frequency Modulated （LFM） signals. The computer simulations demonstrated clearly that the method is effective.