The limitations of fixed kernel functions in the time-frequency analysis of frequency-hopping signals and the improvement of time-frequency resolution based on parameter optimization are analyzed. The changes of entropy measures with different window widths and optional parameters are studied, and compared with the recently proposed Stankovic measure method. The simulation results show that the change law of the same measure is different for different time-frequency representations; only the Flandrin volume normalized entropy gives the optimization results of different time-frequency analyses, so that the amount of information in the time-frequency distribution of frequency-hopping signals can be quantitatively evaluated, which helps to compromise between cross terms and resolution.
You Might Like
Recommended ContentMore
Open source project More
Popular Components
Searched by Users
Just Take a LookMore
Trending Downloads
Trending ArticlesMore