This paper studies the blind source separation (BSS) problem with the assum
ption that the source signals are cyclostationary. Identifiability and sepa
rability criteria based on second-order cyclostationary statistics (SOCS) a
lone are derived. The identifiability condition is used to define an approp
riate contrast function. An iterative algorithm (ATH2) is derived to minimi
ze this contrast function. This algorithm separates the sources even when t
hey do not have distinct cycle frequencies.