Frequency estimation based on variable frequency resolution concept

Frequency estimation is used in various applications, such as wireless communications, acoustic and speech processing, sonar and radar measurements. The classic approach uses the Fourier Transform, which estimates the frequencies at fixed points in frequency domain. This may cause the Fourier analysis to miss the real values of a signal´s frequencies, also known as the picket fence effect. To overcome this issue, different interpolation techniques can be used, resulting in an increase in the estimate process. This paper presents a new algorithm of estimating the frequencies of a signal, using a variable frequency resolution concept.