Fast CS-based pulsar period estimation method without tentative epoch folding and its CRLB

The traditional pulsar period estimation methods have high computational loads as a result of the tentative epoch folding. The tentative epoch folding means that the massive pulsar signals are folded multiple times with different periods. However, due to the low computing power of on-board computer, the tentative epoch folding with high computational load must be avoided. To achieve this objective, we propose a fast compressive sensing-based pulsar period estimation method, which directly estimates the pulsar period according to the distortion degree of the accumulated profile. The compressive sensing (CS) includes three aspects: the dictionary, the measurement matrix and the sparse recovery algorithm. In the proposed method, the pulsar profiles with different distortion degrees comprise the distorted profile dictionary; the low-frequency Fourier transform matrix is developed as the measurement matrix; the super-resolution sparse recovery using the maximal elements is proposed to estimate the pulsar period. Besides, the pulsar period estimation performance against the Cramér-Rao lower bound (CRLB) is studied. And the computational complexity analysis is also provided. The simulation results demonstrate that the fast CS-based pulsar period estimation method approaches the CRLB and its computational time is only 37 ms.