Extended Arnoldi methods for large low-rank Sylvester matrix equations (Q607132)

Two iterative methods for the solution of the low-rank Sylvester equation \(AX+XB+EF^T = 0\) are presented. The proposed methods are projection methods that are based on the extended block Arnoldi (EBA) process and the extended global Arnoldi (EGA) process generating orthonormal bases and \(F\)-orthonormal bases of extended Krylov subspaces. The computation of the residual norm or of an upper bound, without the computation of the approximate solution nor using expensive products with the matrices \(A\) and \(B\), is used to stop the iterations of each algorithm. The method of obtaining a low rank solution of the Sylvester equation in a factored form is also demonstrated. The efficiency and robustness of the proposed methods are evaluated by the help of four numerical examples.