MUSIC法的有限域单向快速频率搜索

Fast Frequency Estimation Way for MUSIC Algorithm

  • 摘要: 快速傅里叶变换(FFT)算法在应用中速度快,精度较低,而MUSIC法精度高,实时性差。该文通过分析两者联合谐波频率估计的可能性,提出利用FFT算法对谐波频率进行预估计,用二分法将频域划分为有限个细小的频率区间,根据可能的有效频率选取有效区间,再通过MUSIC方法在有限的搜索区间进行频率细化,并在细化的过程中采用趋向谱幅值增大的单向搜索。通过仿真研究,该方法能有效提高信号的处理速度,加强MUSIC算法的应用。

     

    Abstract: In the application, fast Fourier transform (FFT) algorithm's speed is fast but its precision is worse whereas MUSIC algorithm has a higher precision but its real time aspect is worse. This paper analyses the characteristics of FFT algorithm and MUSIC algorithm. Using FFT algorithm to pre-estimate a signal harmonic frequency and select some effective frequecy range before using MUSIC algorithm to complete frequency search we can reduce the search area and the processing time of MUSIC algorithm greatly.

     

/

返回文章
返回