Theory, Implementation and Application of Time-Frequency Signal Analysis using the Wigner-Ville Distribution

Abstract

The spectral analysis of non-stationary signals cannot be accomplished by the simple use of classical time domain representations such as correlation methods or frequency domain representations based on the Fourier Transform. For such signals, a concept of Time-Frequency Distribution (TFD) has been defined which takes jointly into account the variables of time and frequency. One of these TFD's, the Wigner-Ville distribution has been shown to possess the most interesting properties for use as a tool for Time-Frequency Signal Analysis. Its properties are reviewed, and it is shown how they can be useful for a practical analysis. The computational aspects for efficiently implementing the method are discussed, and an example of application to the analysis of transient signals is presented.