Processing of Second Order Statistics via Wavelet Transforms

Abstract

Research was conducted to evaluate the feasibility of wavelet transforming second order statistics. To identify the modulation type and to extract signal modulation parameters of frequency hopped signals wavelet functions are applied to the two-dimensional instantaneous correlation function. Parameters of interest are switching times and hop frequencies. The characteristics of the wavelet transform of the correlation function is derived. The wavelet transform is applied along the delay and the time axes. Processing along the delay axis allows the identification of frequency hopped signals and estimation of frequencies and switching times. Processing along the time axis allows detection of the transition times of frequency and time hopped signals.