Optimized Gaussian kernel regression for x-ray pulsar signal profile construction

With the continuous advancement of pulsar-based navigation technology, signal processing issues in X-ray pulsar navigation applications have received increasing attention. One of the key factors limiting the improvement of navigation accuracy is the low signal-to-noise ratio (SNR) of epoch-folded profiles. Pulsar profiles contain many important characteristic features, such as pulse shape symmetry and multi-pulse structures. To facilitate further study, it is necessary to reduce noise and obtain high-SNR pulsar profiles. This study addresses this issue by proposing a method for high-precision pulsar profile construction based on optimized Gaussian kernel regression. The constructed profiles are evaluated and analyzed in terms of profile quality and phase information estimation based on the profile. The results show that the SNR of the epoch-folded profiles after denoising is significantly better than that of the original profiles, providing an important reference for signal processing in X-ray pulsar navigation applications.