Abstract: In order to solve the problem that dense frequency signal can not be measured accurately under the condition of asynchronous sampling, a detection method of dense frequency signal was proposed. The dense frequency signal model was established based on the discrete Fourier transform. And the single channel observation signal was extended periodically to form a multi-channel signal. Then, variable principal component analysis was used to remove the correlation. The dense frequency signal model was solved based on the maximum likelihood blind source separation method. Finally, the parameters of each frequency component were corrected according to the scale coefficient. Simulation and experimental results show that the proposed method has high accuracy and certain anti-noise ability.