Comments on 'A blind signal separation method for multiuser communications'

This correspondence first shows that the global convergence analysis of the method proposed in the above paper is incomplete. Then we provide a counter example to show that the sufficient condition for global convergence is incorrect.