SIMBAD references

X-ray pulsar signals commonly have a very low amplitude and suffer from high observation noise and statistical correlation interference, which seriously degrade the signal detection performance given finite observation times. Here, noise refers to the uncertainty distribution of the photon Poisson statistics, and interference refers to time-dependent broadband or narrowband signals, which may originate from some artificial interference source or other pulsars. To address the above problems, this paper proposes a feature representation method for pulsar signal detection based on high-order cyclostationarity theory. First, we establish a photon signal model of an X-ray detector based on the cyclostationary characteristics of pulsar signals, considering period jitter, correlation interference, and Gaussian noise in the pulsar emission process. By introducing cyclic statistics, we use a direct estimation method for the high-order spectrum, including a nonuniform sampling strategy, to accurately extract signal features and effectively suppress correlation interference and noise. Finally, we use simulation data and Rossi X-ray Timing Explorer observation data to verify the proposed method. The results show that the proposed method is a promising feature modeling solution that is superior to traditional one-dimensional spectra and epoch folding in counteracting interference and noise. While such feature modeling is particularly useful for applications such as pulsar navigation, we believe that this method has general promise for a wide range of pulsar-related research.