СПОСІБ АВТОМАТИЧНОГО ВИЗНАЧЕННЯ НЕСУЧОЇ ЧАСТОТИ КОРОТКОТРИВАЛИХ СИГНАЛІВ ІЗ ЧАСТОТНОЮ МАНІПУЛЯЦІЄЮ

Abstract

Method for determination of the carrier frequency of short-time frequency-shift keying signals based on searching for dominant harmonics whose frequency difference is close to the subcarrier spacing frequency is proposed in the article. In telecommunication systems with burst transmissions, the symbol distribution law of the modulating sequence is not always uniform, which under a limited signal duration leads to an increase in the errors in determining the frequency-shift keying parameters. The increase in errors is due to distortion of the amplitude-frequency spectrum of the signal and a decrease in the probability of the correct identification of the subcarrier oscillations harmonics. To improve the accuracy of determining the carrier frequency proposed method uses a priori information about the value of the spacing frequency of the subcarrier oscillations and the multiplicity of frequency-shift keying. The developed approach consists in obtaining the amplitude-frequency spectrum of the signal, calculating the search threshold of dominant harmonics, determining their frequencies which correspond to subcarrier oscillations, and calculating the carrier frequency as their arithmetic mean value. To calculate the signal spectrum was used the modified Welch periodogram method with has a low dispersion of spectral estimates. The harmonic search threshold is determined automatically based on the statistical characteristics of the amplitude-frequency spectrum. Among harmonics with amplitudes greater than the threshold value, those are determined that the frequency difference between which is closest to the subcarrier spacing frequency. The simulations performed in the MATLAB software environment showed that the error in determining the carrier frequency was halved compared to the classical approach under signal-to-noise ratio from -15 dB, and the developed method was efficient if the symbol appearance probability were from 0,5 to 0,8.