Some convergence conditions for a class of parallel decomposition-type linear relaxation methods (Q1349133): Difference between revisions

The author describes a four-parameter parallel decomposition-type accelerated overrelaxation algorithm for numerical solution of large sparse systems of linear equations on SIMD multiprocessor systems. Convergence is firstly proved for the class of block strictly diagonally dominant matrices and then extended to the whole class of block H-matrices, under appropriate and easy to verify assumptions involving the four relaxation parameters.