A New Time/frequency Technique for Transionospheric Signal Analysis

A new time/frequency technique is developed to characterize dispersed signals in the joint time/scale domain. Specifically, we are interested in pulses that become dispersed by transmission through the ionosphere and are manifested as nonlinear chirps. Due to low signal-to-noise ratios, it is difficult to isolate these signals in the time or frequency domain. Due to computational complexity, it is not practical to examine and detect the signals in real time using Fourier techniques. Standard wavelet methods have been shown to be successful in transionospheric signal analysis [1], but we find they are not the best way to represent the signals of interest. A new technique is developed here, based on wavelets, tailored to capture the characteristics of a nonlinear chirp pulse. Actual satellite data are examined, using the new technique as well as standard Fourier and wavelet methods. Experimental results demonstrate that the new method is a useful tool in isolating the nonlinear chirp pulses at very low signal-to-noise ratios.