Windowing Method for DFT-Based Frequency Estimation of Multiple Complex Exponential Signals

This paper proposes a windowing method for the conventional DFT-based frequency estimation algorithm, restricted initially to rectangular window functions. The proposed method applies a window function to multiple exponential signals to mitigate interference in fractional frequency estimation. A bias correction factor is designed based on the DFT coefficient combination in frequency estimation. Numerical simulations demonstrate that the proposed method maintains consistent estimation performance for various interference levels. Furhter, the results show that Hamming window maintains estimation performance under high interference, while Chebyshev and Blackman windows show sensitivity to the interference levels. This paper provides valuable insights for selecting appropriate window functions in the frequency estimation of multiple exponential signals, which offers a practical and efficient solution for applications in wireless communication.