Title: Blind source separation using second-order cyclostationary statistics
Abstract: Blind source separation (BSS) is a fundamental problem in signal processing. This paper introduces new source separation techniques relying only on cyclostationary second-order statistics (SOS). The identifiability condition and several SOS-based separation criteria are presented for the cyclostationary sources. The corresponding algorithms are developed to achieve these criteria. Simulation results illustrate the performances of the proposed techniques.