Scholarlink

<p>Accurate and fast frequency estimation of complex sinusoidal signal is a fundamental subject in statistical signal processing and plays an important role in many applications, such as communication, radar, and instrumentation. This paper proposes a general frequency estimation method for complex sinusoid based on interpolation of discrete time Fourier transform (DTFT) spectral lines situated at arbitrary positions in the main lobe of the frequency spectrum. Firstly, discrete Fourier transform (DFT) is performed on the sinusoid, and the rough frequency estimation is performed by searching for the position of the DFT spectral line with the maximum amplitude. Then different from all the existing DFT-based algorithms, the maximum DFT spectral line and two DTFT spectral lines situated at arbitrary positions within the main lobe are used for the fine estimation. Simulation results show that the mean squared error of the proposed estimator is closer to the Cramer-Rao lower bound (CRLB) than the competitive estimators without increasing the computational complexity.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Sinusoidal Frequency Estimation Based on Interpolation of DTFT Spectral Lines at Arbitrary Positions in the Main Lobe

  • 期刊论文

  • 发表日期：

    2025年6月23日

查看全文
  • Mingjie Li,
  • Lei Fan,
  • Fei Teng,
  • Shan Wang,
  • Xinrui Zhao,
  • Jiyu Jin

摘要

Accurate and fast frequency estimation of complex sinusoidal signal is a fundamental subject in statistical signal processing and plays an important role in many applications, such as communication, radar, and instrumentation. This paper proposes a general frequency estimation method for complex sinusoid based on interpolation of discrete time Fourier transform (DTFT) spectral lines situated at arbitrary positions in the main lobe of the frequency spectrum. Firstly, discrete Fourier transform (DFT) is performed on the sinusoid, and the rough frequency estimation is performed by searching for the position of the DFT spectral line with the maximum amplitude. Then different from all the existing DFT-based algorithms, the maximum DFT spectral line and two DTFT spectral lines situated at arbitrary positions within the main lobe are used for the fine estimation. Simulation results show that the mean squared error of the proposed estimator is closer to the Cramer-Rao lower bound (CRLB) than the competitive estimators without increasing the computational complexity.