Abstract:T. Saito established a ramification theory for ring extensions locally of complete intersection. We show that for a Henselian valuation ring $A$ with field of fractions $K$ and for a finite Galois extension $L$ of $K$, the integral closure $B$ of $A$ in $L$ is a filtered union of subrings of $B$ which are of complete intersection over $A$. By this, we can obtain a ramification theory of Henselian valuation rings as the limit of the ramification theory of Saito. Our theory generalizes the ramification theory of complete discrete valuation rings of Abbes-Saito. We study "defect extensions" which are not treated in these previous works.
What problem does this paper attempt to address?
The main problem that this paper attempts to solve is to establish the upper ramification group theory in arbitrary valuation rings. Specifically, the authors strive to generalize the ramification theory of Saito and Abbes - Saito et al. in complete intersection ring extensions to be applicable to Henselian valuation rings. They achieve this by proving that the integral closure can be expressed as a filtered union of finite flat and complete intersection rings, and further study "defect extensions", which were not covered in previous work.
### Main Objectives
1. **Generalize the Ramification Theory**: Generalize the existing ramification theory from discrete valuation rings to more general Henselian valuation rings.
2. **Handle Defect Extensions**: Explore the non - trivial defect situations that may occur when the residue field characteristic is \( p>0 \), that is, both the residue field of the extended image and the valuation group extension may be trivial.
### Specific Problem Description
- **Define Upper Ramification Groups**: For a Henselian valuation ring \( A \) and a finite Galois extension \( L \) of its quotient field \( K \), define the upper ramification groups \( G_I^{\log} \) and \( G_I^{n\log} \) in the integral closure \( B \).
- **Compatibility and Generalization**: Ensure that these new definitions of ramification groups are compatible with the existing Abbes - Saito theory on discrete valuation rings and can be generalized to a broader class of rings.
- **Study of Defect Extensions**: Pay special attention to those "defect extensions" that were not dealt with in previous work, that is, when the residue field characteristic of \( A \) is \( p > 0 \), there exists a cyclic extension \( L/K \) of order \( p \), where both the residue field and the valuation group extensions are trivial.
### Main Contributions of the Paper
- **Theoretical Generalization**: By introducing new mathematical tools and techniques, such as log - smooth extensions and adic spaces, the ramification theory has been successfully generalized to a broader class of rings.
- **New Results**: Some important theorems have been proved, such as Theorem 1.1, which shows how to capture the important ideal \( H \) with the upper ramification groups and can work effectively even in the presence of defect extensions.
In conclusion, this paper fills the gaps in the existing theory in dealing with broader classes of rings and special extension cases by in - depth research and generalization of the ramification theory, providing a solid foundation for further research in related fields.