Adaptive pulsar time delay estimation using wavelet-based RLS

Abstract

Pulsar time delay needs to be quickly and accurately estimated. In this study, a novel pulsar time delay estimation (PTDE) method based on the wavelet transform and the recursive least squares (RLS) is proposed because of its good robustness and fast convergence of the RLS algorithm. However, the PTDE based on the RLS algorithm (RLSPTDE) leads to high complexity. To address this problem, we initially perform multilevel Haar wavelet transform on the pulsar profile to obtain low-frequency signal. Then, according to the low frequency signal, the adaptive filter based on the RLS algorithm is used to obtain the pre-estimated time delay value. Finally, according to the pre-estimated time delay value, we use the adaptive filter based on the RLS algorithm to perform locally accurate time delay estimation. Certain benefits can be derived from using adaptive filters, especially when underlying parameters, such as signal statistics, are unknown or have changed over time. The theoretical analysis and experimental results show that the proposed method can reduce computational complexity and achieve higher PTDE accuracy than the RLSPTDE.