On Milnor $K$-theory in the imperfect residue case and applications to period-index problems

Basic Information

• Paper ID: 2510.03603
• Title: On Milnor $K$-theory in the imperfect residue case and applications to period-index problems
• Author: Srinivasan Srimathy (Tata Institute of Fundamental Research, Mumbai)
• Classification: math.NT (Number Theory), math.AG (Algebraic Geometry), math.KT (K-theory), math.RA (Rings and Algebras)
• Publication Date: October 2025 (arXiv preprint, v2)
• Paper Link: https://arxiv.org/abs/2510.03603

Abstract

This paper studies $(0,p)$-mixed characteristic complete discrete valuation fields $K$ and defines a class of finite field extensions called "pseudo-perfect extensions," such that for all $p \neq 2$, the natural restriction map on mod-$p$ Milnor $K$-groups vanishes. This implies that pseudo-perfect extensions split every element in $H^i(K,\mu_p^{\otimes i-1})$, thereby providing period-index bounds for Brauer classes and higher cohomology classes. As a corollary, the author proves the Bhaskhar-Haase conjecture: the Brauer $p$-dimension of $K$ does not exceed $n+1$, where $n$ is the $p$-rank of the residue field.

Research Background and Motivation

Core Problem

The core problem addressed in this paper is the period-index problem, a classical problem in algebra:

1. For an element $\alpha$ in the Brauer group $\operatorname{Br}(F)$, the relationship between its period $\operatorname{per}(\alpha)$ (the order in the Brauer group) and its index $\operatorname{ind}(\alpha)$ (the greatest common divisor of degrees of splitting fields)
2. Finding the minimal integer $\ell$ such that $\operatorname{ind}(\alpha) | \operatorname{per}(\alpha)^\ell$ for all $\alpha$

Significance

1. Theoretical importance: Brauer dimension is a crucial invariant of fields, characterizing the structure of central simple algebras
2. Technical challenges: When the residue field is imperfect (characteristic $p$ with finite $p$-rank), the problem becomes extremely complex
3. Applied value: Results apply to arithmetic geometry, algebraic $K$-theory, and other fields

Limitations of Existing Methods

1. Perfect residue field case: Relatively good results exist (Harbater-Hartmann-Krashen, etc.)
2. Imperfect residue field: Parimala-Suresh provided bounds, but they are not sharp
3. Technical difficulties: Lack of effective tools for handling Milnor $K$-theory in the mixed characteristic case

Core Contributions

1. Introduction of pseudo-perfect extensions: Defines a new class of field extensions, analogous to perfect closures in the equicharacteristic case
2. Proof of main theorem: For $p \neq 2$, the restriction map on mod-$p$ Milnor $K$-groups over pseudo-perfect extensions vanishes
3. Resolution of Bhaskhar-Haase conjecture: Proves the optimal upper bound $R_p(\kappa)+1$ for Brauer $p$-dimension
4. Improved semi-global field bounds: Provides sharper Brauer $p$-dimension bounds than Parimala-Suresh
5. Unified framework: Provides a unified approach for handling both mixed characteristic and equicharacteristic cases

Detailed Methods

Core Concept Definitions

Pseudo-rank and pseudo-basis

For the fraction field $K$ of a complete regular local ring $R$, define:

• Pseudo-rank: $R_{ps}(K) := R_p(\kappa) + \dim R$
• Pseudo-basis: $\Lambda = \tilde{B} \cup \Pi$, where $\tilde{B}$ is a lift of a $p$-basis of the residue field $\kappa$, and $\Pi$ is a regular parameter system

Pseudo-perfect extensions

An $\ell$-level pseudo-perfect extension is defined as: $L \cong K(\Lambda^{1/p^\ell})$

This mimics the construction of perfect closures in the equicharacteristic case.

Main Technical Tools

Kato filtration theory

Utilizing Kato's construction of filtrations on unit groups of complete discrete valuation fields: $U^i_K = \{1 + \pi^i O_K^\times\}$

For $e'_K = e_K p(p-1)^{-1}$, we have $U^{e'_K+1}_K \subseteq (O_K^\times)^p$.

Key Lemma (Key Lemma 6.8)

Let $w = \{w_1,\ldots,w_i\} \in K_i(K)/p$, where each $w_j \in O_K^\times$ is a finite sum of $p$-th powers in $O_K$. Then:

• When $p \neq 2$: $w = 0$
• When $p = 2$ and $i \geq 3$: $w = 0$
• When $p = 2$ and $i = 2$: $w = 0$ under certain conditions

Proof Strategy

Main theorem proof outline

1. Reduction to units: By multiplying by appropriate powers of uniformizing elements, reduce symbols to elements of unit groups
2. Expansion representation: Use lifts of $p$-bases to represent each unit in a specific form
3. Application of key lemma: In pseudo-perfect extensions, the expansion becomes a sum of $p$-th powers, and the key lemma yields vanishing

Technical Innovation

1. Symbol manipulation: Clever use of symbol relations in Milnor $K$-theory
2. Filtration techniques: Combining Kato filtrations with unit group structure
3. Cohomological correspondence: Connecting $K$-theory and Galois cohomology via norm residue isomorphism

Experimental Setup

Theoretical Verification

This is primarily a theoretical work; "experiments" manifest as:

1. Explicit example construction:
   • Section 7 constructs explicit examples where the restriction map is nonzero for $p=2$
   • Uses $\kappa = \mathbb{F}_2(a,b)$ and $K$ as a characteristic zero complete discrete valuation field
2. Boundary case analysis:
   • Verifies that the $p=2$ case requires the additional condition $H^2_{\text{ét}}(\kappa,\mathbb{Z}/2(1)) = 0$
   • Analyzes the necessity of this condition

Application Verification

The validity of the theory is verified through:

1. Proof of Bhaskhar-Haase conjecture
2. Improvement of semi-global field bounds
3. Unified bounds for higher cohomology

Experimental Results

Main Results

Theorem 4.1 (Main Theorem)

Let $K$ be a characteristic zero complete discrete valuation field with residue field $\kappa$ of characteristic $p$. Then for any $L \in PP(K)$: $\operatorname{Res}^i_{L/K}: K_i(K)/p \to K_i(L)/p$

• vanishes for $i \geq 3$
• vanishes for $i = 2$ when $p \neq 2$ or $H^2_{\text{ét}}(\kappa,\mathbb{Z}/2(1)) = 0$

Corollary 4.5 (Bhaskhar-Haase Conjecture)

$R_p(\kappa) \leq \operatorname{Br}_p\dim(K) \leq R_{ps}(K) = R_p(\kappa) + 1$

Theorem 4.9 (Semi-global Field Improved Bound)

For semi-global fields $F$: $\operatorname{Br}_p\dim(F) \leq R_{ps}(K) + 1$

This is sharper than the Parimala-Suresh result.

Technical Achievements

1. Uniformity: Bounds for mixed characteristic and equicharacteristic cases coincide
2. Optimality: Achieves the conjectured upper bound
3. Constructivity: Provides explicit constructions of splitting fields

Boundary Case Analysis

Peculiarity of $p=2$

Theorem 7.2 shows that when $p=2$ and $H^2_{\text{ét}}(\kappa,\mathbb{Z}/2(1)) \neq 0$, the restriction map may be nonzero, demonstrating the necessity of the additional condition.

Related Work

Historical Development

1. Classical results: Albert's theorem (Brauer $p$-dimension bounds for characteristic $p$ fields)
2. Modern developments:
   • Harbater-Hartmann-Krashen: Residue characteristic coprime to $p$
   • Parimala-Suresh: Bounds when residue characteristic equals $p$
   • Bhaskhar-Haase: Conjecture on optimal bounds

Technical Connections

1. Milnor $K$-theory: Bloch-Kato theorem, norm residue isomorphism
2. Complete discrete valuation field theory: Kato's filtration theory
3. Patching techniques: Harbater-Hartmann-Krashen methods

Advantages of This Paper

1. Unified framework: Unified approach for mixed characteristic and equicharacteristic cases
2. Optimal bounds: Achieves the conjectured optimal upper bound
3. Constructive methods: Provides explicit field extension constructions

Conclusions and Discussion

Main Conclusions

1. Pseudo-perfect extension theory: Successfully extends the perfect closure theory from equicharacteristic to mixed characteristic
2. Optimal Brauer bounds: Resolves the Bhaskhar-Haase conjecture with optimal upper bounds
3. Technical breakthrough: Develops new techniques for handling imperfect residue fields

Limitations

1. $p=2$ case: Requires additional technical conditions; more complex handling
2. Regularity assumptions: Requires complete regular local ring assumptions
3. Finite $p$-rank: Requires residue field to have finite $p$-rank

Future Directions

1. Problem 7.4: Complete resolution of the conjecture for $p=2$
2. Higher-dimensional generalizations: Extension to more general arithmetic geometry settings
3. Computational aspects: Development of effective algorithmic implementations

In-Depth Evaluation

Strengths

Theoretical Innovation

1. Conceptual innovation: The pseudo-perfect extension concept elegantly unifies different characteristic cases
2. Technical breakthrough: Key Lemma 6.8 reveals deep properties of $p$-th power sums in Milnor $K$-theory
3. Unified framework: Provides a unified approach to period-index problems

Mathematical Depth

1. Multi-disciplinary fusion: Cleverly combines algebraic $K$-theory, Galois cohomology, and arithmetic geometry
2. Technical sophistication: Deep application of Kato filtration theory demonstrates high technical mastery
3. Optimal results: Achieves the conjectured optimal bound with significant theoretical value

Proof Rigor

1. Clear logic: Proof structure is clear and steps are rigorous
2. Complete details: Technical details are handled appropriately with sufficient boundary case analysis
3. Counterexample construction: Section 7's counterexample construction demonstrates deep understanding of the problem's essence

Weaknesses

Technical Limitations

1. $p=2$ complexity: This case requires additional conditions; theory is not fully unified
2. Strong assumptions: Complete and regular assumptions limit the scope of applications
3. Limited constructivity: Some proofs rely on existence arguments

Presentation Issues

1. High technical threshold: Requires deep background in algebraic geometry and $K$-theory
2. Complex notation: Abundant technical notation may affect readability

Impact

Theoretical Contribution

1. Resolution of important conjecture: The resolution of Bhaskhar-Haase conjecture is of milestone significance
2. New research directions: Pseudo-perfect extension theory may inspire further research
3. Technical impact: New techniques applicable to related problems

Practical Value

1. Arithmetic geometry applications: Results apply to elliptic curves, abelian varieties, etc.
2. Cryptographic potential: Brauer groups have potential applications in cryptography
3. Computational significance: Provides theoretical foundation for related computational problems

Applicable Scenarios

1. Algebraic number theory: Study of Brauer group structure on arithmetic fields
2. Arithmetic geometry: Analysis of central simple algebras on algebraic varieties
3. Algebraic $K$-theory: Study of Milnor $K$-group properties
4. Galois cohomology: Analysis of field cohomology group structure

References

The paper cites important literature in the field, including:

• BH20 Bhaskhar-Haase's work on Brauer $p$-dimension
• PS14, PS15 Parimala-Suresh's research on period-index problems
• HHK09 Harbater-Hartmann-Krashen's patching techniques
• BK86 Bloch-Kato's foundational work on $p$-adic étale cohomology
• Kat89 Kato's important results on cohomology of complete discrete valuation fields

This paper represents an important contribution to algebraic number theory and arithmetic geometry. Through the innovative introduction of pseudo-perfect extensions, it successfully resolves the Bhaskhar-Haase conjecture and achieves significant technical breakthroughs. Although the $p=2$ case remains to be perfected, overall this represents important progress in the field.