Chen, L (reprint author), Beijing Normal Univ, Dept Astron, Beijing 100875, Peoples R China.
The method of auto-regressive power density spectrum (ARPDS) is used to analyse the frequency spectrum characteristics of Cyg X-l. Based on our results, we conclude that the power spectrum can be modelled in the same way for the transition and soft states by a broken power law plus noise. There is a wide noise component (< 3 Hz) and a QPO (4-12 Hz) in each of the transition states, while in the soft state, there is no significant QPO, and its PDS can be characterized perfectly by a broken power law. The frequency cutoff in each state is quite stable, and its evolution is consistent with the qualitative features predicted by the ADAF model. Our study shows that the method is effective for searching and identifying the frequency components in PDS.