METHOD FOR ESTIMATING THE DOPPLER FACTOR OF A BROADBAND SIGNAL UNDER CONDITIONS OF MULTIPLICATIVE INTERFERENCE WITH UNKNOWN PROBABILITY DISTRIBUTION

Background. The paper presented a method for estimating the Doppler factor of a received broadband pulse signal distorted by multiplicative interference with an a priori uncertain probability distribution. This paper was considered a case where the multiplicative interference in the observation has a Gaussian distribution Special attention is paid to the logistic distribution of interference, which is characterized by "heavier tails" of the probability density distribution. Materials and Methods. The Doppler frequency of the signal is estimated by minimizing the objective function using the SG statistics of the observation. There was an investigated estimation quality of Doppler frequency by statistical modeling of the PSK signal, interference, and the objective function minimization algorithm. There was completed a comparative analysis of estimations of the signal`s Doppler frequency obtained by the proposed nonparametric method and the maximum likelihood method for different interference parameters. Results and Discussion. It was shown that the objective functions obtained for multiplicative interference with Gaussian and logistic distributions have clear extrema, which eliminates the possibility of ambiguous estimation of the signal`s Doppler frequency. Histograms of the distribution of the signal`s Doppler frequency estimates by the proposed method show that, regardless of the considered laws of the probability density distribution of multiplicative interference, the estimates are concentrated in the vicinity of the actual value of the signal`s Doppler frequency. The distribution of estimates using the maximum likelihood method doesn`t have a clear extremum in the researched frequency interval.The modeling results showed that in conditions of signal distortion by multiplicative interference, the error in estimating the signal`s Doppler frequency, which is obtained by maximizing the cross-correlation function of the observation and the expected signal, is significantly greater than when using the objective function based on SG statistics. Conclusion. The defined parameters of SG statistics allow obtaining a qualitative estimate of the signal`s Doppler frequency, independent of the law`s probability density distribution of multiplicative interference, which indicates the effectiveness of the proposed method.