An iterative method applied to nonsymmetric linear systems (Q1922227)

A two-stage Gauss-Seidel method for solving systems of linear equations with coefficient matrices being the identity plus a skew-symmetric part is reported. The most popular in this case, the extrapolated iterative methods, need an estimation of the optimum extrapolation parameter, which is often a difficult part of the problem. The iterative method proposed in this paper works without a parameter estimation. The idea is to apply two times the classical Gauss-Seidel method. The convergence theorems for this two-stage Gauss-Seidel method are proven. Two numerical examples are added to present the behaviour of the new algorithm. The results show that the number of iterations needed of the two-stage Gauss-Seidel method is quite smaller than of the classical variant and does not depend on the order of the coefficient matrix.