Performance analysis of iterative algorithms for sylvester equations

We consider the convergence performance of the iterative algorithms proposed by Ding and Chen [2], [3] for the solution of coupled matrix equations. A stiffness property is given to describe equations where these algorithms converge only very slowly, and a stochastic analysis shows that very slow convergence is a generic feature of these algorithms. Lastly we consider the recently introduced modifications to these algorithms proposed by Zhou et al [9], [10] and discuss their potential to improve the convergence speed for such stiff equations.