COMBINED TIME-FREQUENCY KERNELS AND POLYNOMIAL PHASE MODELING FOR SIGNAL CHARACTERIZATION

This paper introduces some tools to characterize the phase behav-ior of non stationary signals. Starting from various kernels used for Time-Frequency representation, we apply polynomial phase model-ing to accurately estimate the Instantaneous Frequency Law (IFL) of the signal. The method is also extended for higher order phase derivatives. We theoretically prove the relevancy of our undertaking and provide some examples to show the accuracy and efficiency of the method. 1