Localized time-frequency analysis of signals using wavelet transform

Abstract

Simulated sinusoidal signals are analyzed locally in time and frequency domains using the dilation and translation properties of wavelet transform with respect to Daubechies wavelet. The method employed makes use of the pyramidal algorithm to decompose a signal into its resolution levels. For an N-datapoint signal, N wavelet coefficients need to be calculated. These signals are then reconstructed using the inverse wavelet transform.