Graph Signal Processing Meets Blind Source Separation

In graph signal processing (GSP), prior information on the dependencies in is collected a which then used when or analyzing signal. Blind source separation (BSS) techniques have been developed and analyzed different domains, but for signals research BSS still its infancy. this paper, gap filled with two contributions. First, nonparametric method, relevant to GSP framework, refined, Cram\'{e}r-Rao bound (CRB) mixing unmixing matrix estimators case of Gaussian moving average derived, studying achievability CRB, new parametric method introduced. Second, we also consider non-Gaussian methods are proposed. Identifiability conditions show that utilizing both structure non-Gaussianity provides more robust approach than based only either non-Gaussianity. It demonstrated by numerical study proposed efficient separating signals.