Time-Frequency Analysis of Nonstationary Optical Signals Using Husimi Type Function

Abstract

Time-frequency analysis is important to identify the localized information of a non-stationary signal in the time and frequency domains simultaneously. There are few different time-frequency analysis methods available with their own specialty and suitability. In quantum mechanics Husimi function of any quantum mechanical state arises naturally whenever the simultaneous measurement of both coordinate and momentum is performed on this state with maximal accuracy allowed by the Heisenberg uncertainty relations. The Husimi function is a probability distribution for the statistics of simultaneous unsharp measurement of both coordinate and momentum. In general, like the Wigner function, the Husimi function can be defined in the space of any pair of conjugate variables. In particular, in the studies of signal processing, this space is the time-frequency space. In the present work we consider the Husimi function in this space, and apply it to analyse the multicomponent signals. The time-frequency representations of the simulated signals by using the Husimi distribution clearly show the frequency features along the time axis. The results are encouraging and indicate that, like in corresponding analogous problems in quantum mechanics, the Husimi distribution approach in the time-frequency analysis for non-stationary optical signals may provide some insights which are not so easily obtained in other, more spread approaches.