Adaptive Window Width Selection Algorithm for Gabor Transform Based on Improved Shannon Entropy
Dailei Zhang, Huabin Wang, Zhou Jian, Tao Liang · 2020
Non-stationary signal is a common signal in reality, but the existing single window discrete Gabor transform algorithm can not effectively display all signal components or the time-frequency accuracy is not high, so this paper proposes an improved algorithm based on Shannon entropy adaptive selection analysis window width. Firstly, the Gabor transformation coefficient is normalized, and Shannon entropy is obtained according to the transformation coefficient. Then, the value range of Shannon entropy is improved, and the time frequency aggregation degree is determined according to the value of Shannon entropy. Finally, the Shannon entropy of all window functions from the minimum window width to the maximum window width is calculated, and the window width corresponding to the maximum Shannon entropy is the optimal window width. This algorithm can not only select an appropriate window width to display all components when processing multi-component signals, but also improve the time-frequency accuracy. Even for multi-component signals with different frequency characteristics, it can choose a better analysis window and has a good anti-noise capability.