Comments on blind beamforming for multiple non-gaussian signals and the constant-modulus algorithm

In a recent paper, Sheinvald compared a cumulant matching criterion to thre e simplified criteria that he claimed to be equivalent. We show that tno si mplified criteria can be viewed as direct consequences of a result already published and that their equivalence must be revised. We also give addition al references of closely related works that have been omitted, some of them presenting Jacobi-like source separation algorithms, presented as original in by Sheinvald.