Abstract:The solution of systems of linear(ized) equations lies at the heart of many problems in Scientific Computing. In particular for systems of large dimension, iterative methods are a primary approach. Stationary iterative methods are generally based on a matrix splitting, whereas for polynomial iterative methods such as Krylov subspace iteration, the splitting matrix is the preconditioner. The smoother in a multigrid method is generally a stationary or polynomial iteration. Here we consider real symmetric indefinite and complex Hermitian indefinite coefficient matrices and prove that no splitting matrix can lead to a contractive stationary iteration unless the inertia is exactly preserved. This has consequences for preconditioning for indefinite systems and smoothing for multigrid as we further describe.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is about the effectiveness of splitting and preconditioning methods for symmetric indefinite matrices in solving linear equations. Specifically, the author explores how to select appropriate splitting matrices or preconditioning matrices when using iterative methods to solve large - scale linear equations, in order to ensure the convergence and effectiveness of the iterative process.
### Main Problem Summary:
1. **Splitting Matrix and Contraction**:
- For real symmetric indefinite matrices and complex Hermitian indefinite matrices, the author proves that unless the inertia (i.e., the number of positive, zero, and negative eigenvalues remains unchanged) is maintained, a fixed - point iteration of contraction cannot be achieved through the splitting matrix.
- This means that if the splitting matrix cannot fully maintain the inertia of the original matrix, the iterative process will not be contractive, which may lead to non - convergence or slow convergence.
2. **Preconditioning and Multigrid Smoothing**:
- Preconditioning techniques are very important in solving indefinite systems, especially when used as smoothers in the multigrid method.
- The author points out that when the inertia of the preconditioning matrix is inconsistent with that of the coefficient matrix, convergence cannot be guaranteed even with the standard Chebyshev polynomial acceleration.
- Therefore, in order to ensure the effectiveness of the iterative method, a preconditioning matrix that can maintain inertia must be selected.
3. **Application of Vanka - type Iteration**:
- Vanka - type iteration is widely used in solving indefinite saddle - point problems such as Stokes equations, but its convergence analysis is still limited.
- The author emphasizes the importance of maintaining inertia in these cases and points out that even a successful Vanka - type iteration must meet specific conditions, such as the clustering of eigenvalues.
### Key Conclusions:
- **Selection of Preconditioning Matrix**: For indefinite systems, the preconditioning matrix should maintain the same inertia as the coefficient matrix as much as possible, otherwise the iterative method may not converge.
- **Multigrid Smoothing**: When using indefinite matrix splitting as a multigrid smoother, if the inertia is not fully maintained, the smoother cannot be a contractive iteration.
- **Combination of Theory and Application**: Although maintaining inertia is theoretically necessary, it may be challenging to find an appropriate preconditioning matrix in practical applications, especially in wave problems (such as Helmholtz equations).
Through these analyses, the author provides important theoretical guidance for the iterative solution of uncertain systems and points out the direction for future research.