Pulsar profile denoising using kernel regression based on maximum correntropy criterion

A pulsar profile denoising method using kernel regression based on maximum correntropy criterion is proposed. This method uses the kernel regression to reduce the human visual inspection inescapable in the current profile denoising methods and the reliance of the prior knowledge on the profile of interest. In order to cope with the non-Gaussian case that is common in a real application, the maximum correntropy criterion is introduced into the kernel regression to resist the impact of non-Gaussian noise. The performance of the prosed method is verified via simulation and real data. The results have shown that the proposed method outperforms the current signal denoising methods in a non-Gaussian environment and is readily to be applied.