ACCURACY ESTIMATION OF MEASURING THE INSTANTANEOUS FREQUENCIES AND PHASES OF COMPONENTS OF THE MULTICOMPONENT NON-STATIONARY SIGNAL

Measurements of instantaneous frequencies and phases of non-stationary components of the multicomponent signals are actual in processing medical, radar, sonar, seismic, vibration, and speech signals. When we measure the parameters of multi-component non-stationary signals, it is a need to measure the instantaneous frequency and current phase of several non-stationary frequency-modulated signal components. In this article, estimates of the measurement accuracy of the specified parameters of the non-stationary signal components were obtained using the method of simulation modeling. Measurement methods based on the short-time Fourier transform and the proposed method with local adaptive matching filtering of signal components in the frequency domain in the time-sliding observation window were studied. Adjustment of matched filters is performed separately for each component of the signal, returning to the time domain and taking into account the filter settings in the previous time window. The simulation method is used for the signal model consisting of three components. The first two components are the frequency modulated harmonic ones and the last is harmonic with constant frequency. The amplitudes of components and additive Gaussian noise variance were variated to study the influence of signal-to-noise ratio on the accuracy of measurements. Estimates of the experimental standard deviation of instantaneous frequencies and current phase measurements were obtained by comparing them with their known dependencies. The frequency laws of non-stationary signal components were chosen to be harmonic, which also allowed us to evaluate the influence of the nonlinearity of the instantaneous frequency change on the measurement accuracy. The results of differential phase measurements between two measurement channels are obtained to illustrate the quality of such ones for multi-component non-stationary signals situations. The results of this study may be useful for the development of imaging methods in inverse synthetic aperture radars (ISAR) and interferometric ISAR.