Enhancing Teager Energy in Time Frequency Transforms with a Concentrated Discrete Window

The time frequency representation (TFR) of a signal offers extensive details regarding the local spectrum at specific time instants. However, a notable drawback of the linear time frequency transforms (TFT) is that it is challenging to obtain high resolution in both time and frequency simultaneously due to the Heisenberg's ambiguity principle. This limitation hampers the effectiveness of linear TFTs in accurately determining the temporal components of a signal, and detecting finite duration events. To address this issue, a modified version of the linear TFTs, incorporating the proposed Teager energy concentrated discrete window (TECDW) based on the optimized discrete Teager energy concentration (DTEC) measure is suggested. This method is especially advantageous for accurately identifying specific finite duration events embedded in noise, and precisely identifying the start and end of such events. The simulated outcomes indicate that the proposed approach can effectively distinguish signal components in the time domain while attaining the highest DTEC in the frequency domain.