TOA Estimation And Identification Of X-ray Pulsar Signals

In astronomical navigation,X-ray pulsar navigation is a new technology that relies on pulsar signal sources and is mainly used for deep space exploration missions.X-ray pulsars have good cycle stability,high flow intensity,and predictable rotational frequency.According to the characteristics of X-ray pulsar signal,two common pulsar signal processing problems are discussed,namely delay estimation problem,denoising and identification problem.The current algorithm still has defects in meeting the requirements of the navigation system,and the efficiency and accuracy cannot meet the requirements of the pulsar navigation system.To this end,research work on pulsar signal processing has been carried out.In order to guarantee the accuracy of TOA estimation in the effective observation time,a TOA estimation algorithm based on Bayesian parameters is proposed.According to the characteristics of the X-ray pulsar signal,the photon sequence of the PSR B0531+21 signal in the EPN database is selected,and the Poisson distribution signal model is established in the frequency domain.According to the flow characteristics of the signal,the probability density function is obtained,and the flow probability likelihood function is derived.Finally,the TOA value in the posterior distribution is solved by the statistical method of Bayesian parameters.The algorithm effectively avoids the disadvantage that the time domain method is limited by the contour sampling frequency.The Bayesian parameter estimation algorithm improves the accuracy of the delay estimation under the condition of low SNR,which provides a new direction for future research on TOA estimation.In order to deal with the denoising and identification of pulsar signals more efficiently,the wavelet entropy processing method is adopted.After the wavelet transform,the pulsar signal is decomposed to obtain the coefficients of each layer,and the wavelet entropy of the highfrequency coefficients of each layer is calculated between the partitions.Because the size of the wavelet entropy can reflect the position of the noise,the noise variance is calculated from the magnitude of the entropy,and the threshold function is constructed to denoise.The wavelet entropy value of each layer coefficient is the characteristic of energy distribution,and the residual value can be used for comparison and identification.In this paper,25 sets of pulsar signal characteristics database are established as an example.The comparison process is quickly completed by comparing the residual values between 21 sets of pulsar signals.For the unknown pulsar signal not in the database,the pulsar name can be included in the database,and the feature information continuously updated in the feature database can also be used as a reference for other research contents.