A stable Richardson iteration method for complex linear systems (Q1100843)

The Richardson iteration method is conceptually simple as well as easy to program and parallelize. This makes the method attractive for the solution of large linear systems of algebraic equations with matrices with complex eigenvalues. We choose the relaxation parameters to be subsets of sets of reciprocals of Fejér points. An ordering of the relaxation parameters is described which ensures that the obtained iterative method is stable and converges asymptotically optimally.