Matrices over maximal orders in cyclic division algebras over Q as sums of squares and cubes

Basic Information

• Paper ID: 2510.13469
• Title: Matrices over maximal orders in cyclic division algebras over Q as sums of squares and cubes
• Authors: S. A. Katre, Deepa Krishnamurthi
• Classification: math.NT (Number Theory)
• Publication Date: October 15, 2025 (arXiv preprint)
• Paper Link: https://arxiv.org/abs/2510.13469

Abstract

This paper investigates the representation of matrices over maximal orders in cyclic division algebras of odd prime degree over the rational numbers Q as sums of squares and cubes. It is known that under certain discriminant conditions, n×n matrices over maximal orders of algebraic number fields can be represented as sums of k-th powers. Wadikar and Katre have proven that every matrix of order greater than 2 over maximal orders in rational quaternion division algebras can be represented as sums of squares and cubes. This paper generalizes this result to cyclic division algebras of odd prime degree, proving that under specific conditions, every matrix of order greater than 2 over maximal orders of these noncommutative rings can be represented as sums of squares and cubes.

Research Background and Motivation

Problem Background

1. Generalization of Waring's Problem: The classical Waring problem studies the representation of positive integers as sums of k-th powers. This paper extends this problem to matrices over noncommutative rings.
2. Matrix Power Sum Representation Theory: Investigating whether matrices can be represented as sums of k-th powers of other matrices is an important problem in algebra, closely related to ring theory and matrix theory.
3. Importance of Cyclic Division Algebras: Cyclic division algebras constitute an important class in noncommutative algebra with widespread applications in algebraic number theory, representation theory, and related fields.

Research Motivation

1. Theoretical Completeness: Generalizing from rational quaternion division algebras to general cyclic division algebras of odd prime degree to perfect the theoretical framework.
2. Technical Challenges: The complexity of noncommutative rings makes the proof techniques more difficult, requiring deep knowledge of algebraic number theory.
3. Application Prospects: Providing theoretical foundations for related problems in algebraic geometry and arithmetic geometry.

Core Contributions

1. Main Theorems: Proves that under specific conditions, matrices over maximal orders in cyclic division algebras of odd prime degree can be represented as sums of squares (Theorem 4) and cubes (Theorem 5).
2. Technical Innovation: Employs Gauss periods, cyclotomic numbers, period polynomials, and other tools combined with canonical generation theory for cyclic division algebras.
3. Characterization of Conditions: Explicitly provides sufficient conditions for matrices to be representable as power sums, including congruence conditions on σ and norms.
4. Proof Methods: Develops systematic methods for handling matrix power sum problems over noncommutative rings.

Detailed Methodology

Theoretical Foundations

Construction of Cyclic Division Algebras

Let n be an odd prime, p = hn + 1, g a primitive root modulo p, and ξ a primitive p-th root of unity. Define the Gauss period: $\eta = \sum_{r=0}^{h-1} \xi^{g^{nr}}$

The cyclic division algebra A has a Q-basis: $\{u^i z_k : i = 0,1,\ldots,n-1; k = 1,2,\ldots,n\}$

satisfying the relations:

• $zu = uz^S$ (where S is a generator of the Galois group)
• $u^n = \sigma \neq 0 \in \mathbb{Q}$

Canonical Generation

According to Ralph Hull's theorem, there exist infinitely many rational primes p satisfying:

• $p \equiv 1 \pmod{n}$ and coprime to σ
• $q_1, q_2, \ldots, q_s$ are n-th non-residues modulo p
• σ is an n-th residue modulo p

Core Proof Strategy

Representation as Sums of Squares (Theorem 4)

Conditions:

1. $\sigma \equiv 1 \pmod{2}$
2. $N_{\mathbb{Q}(\eta)/\mathbb{Q}}(\eta) \equiv 1 \pmod{2}$

Proof Strategy:

1. Prove that all basis elements $\{u^i \eta^{S^j}\}$ belong to $T_2$ (the set of elements representable as sums of squares)
2. Utilize properties of $\eta = \sum_{r=0}^{h-1} \xi^{g^{nr}}$ by analyzing the square properties of $\xi^i$
3. For u, use $u^{n+1} \equiv u \pmod{2}$ (since n is odd)
4. Inductively prove $u^i \eta^{S^j} \in T_2$

Representation as Sums of Cubes (Theorem 5)

Conditions:

1. $3 \nmid \sigma$
2. $3 \nmid N_{\mathbb{Q}(\eta)/\mathbb{Q}}(\eta)$

Proof Strategy:

1. Use $3x \equiv (p+1) \pmod{p}$ has a solution to obtain $\xi \in T_3$
2. Conduct case analysis based on the relationship between n and 3:
   • If $n = 3k+2$: utilize $u^{n+1} \equiv u \pmod{3}$
   • If $n = 3k+1$: requires more refined analysis
3. Inductively prove all basis elements belong to $T_3$

Technical Innovation Points

Comprehensive Application of Mathematical Tools

1. Gauss Period Theory: Exploits properties of period polynomials and cyclotomic numbers
2. Galois Theory: Fully utilizes the automorphism group structure of cyclic extensions
3. Congruence Analysis: Simplifies complex algebraic relations through modular arithmetic

Proof Techniques

1. Basis Induction: Establishes general results by proving properties of all basis elements
2. Case Analysis: Conducts refined case distinctions based on the relationship between n and 2, 3
3. Norm Computation: Utilizes norm properties under Galois group action

Main Results

Theorem 4 (Representation as Sums of Squares)

Statement: Let A be a rational cyclic division algebra of degree n and order n², with canonical generation (σ, Z, S), and let m be the corresponding maximal order. If:

1. $\sigma \equiv 1 \pmod{2}$
2. $N_{\mathbb{Q}(\eta)/\mathbb{Q}}(\eta) \equiv 1 \pmod{2}$

then every matrix in $M_n(m)$ can be represented as a sum of squares of matrices in $M_n(m)$.

Theorem 5 (Representation as Sums of Cubes)

Statement: Under the same setting, if:

1. $3 \nmid \sigma$
2. $3 \nmid N_{\mathbb{Q}(\eta)/\mathbb{Q}}(\eta)$

then every matrix in $M_n(m)$ can be represented as a sum of cubes of matrices in $M_n(m)$.

Concrete Examples

The paper provides detailed calculations for n=3:

• For p=7, obtains the period polynomial $P_3(x) = x^3 + x^2 - 2x - 1$
• Verifies relevant conditions and constructs specific cyclic division algebras

Related Work

Historical Development

1. Griffin-Krusemeyer: First studied representation of matrices over commutative rings as sums of squares
2. Katre-Garge: Generalized to the case of algebraic number fields
3. Wadikar-Katre: Handled the case of quaternion division algebras

Contribution of This Paper

This paper generalizes results from quaternion division algebras (degree 2) to general cyclic division algebras of odd prime degree, representing an important theoretical advance.

Conclusions and Discussion

Main Conclusions

1. Under appropriate conditions, matrices over maximal orders in cyclic division algebras of odd prime degree can be represented as sums of squares and cubes
2. The conditions primarily involve congruence properties of the generator σ and the Gauss period norm
3. The proof methods are general and may be extended to other cases

Limitations

1. Condition Restrictions: Requires satisfaction of specific congruence conditions; not all cases are covered
2. Degree Restrictions: Only addresses odd prime degree cases
3. Open Problems: The authors explicitly indicate that the problem remains open when conditions are not satisfied

Future Directions

1. Remove restrictions on congruence conditions
2. Extend to even or composite degree cases
3. Investigate representation problems for higher powers

In-Depth Evaluation

Strengths

1. Theoretical Depth: Synthesizes profound mathematical tools from algebraic number theory, Galois theory, and cyclotomic theory
2. Rigorous Proofs: Provides complete proofs with clear logical structure
3. Generalization Significance: Generalizes from special cases to general cases with important theoretical value
4. Technical Innovation: Develops new methods for handling matrix power sum problems over noncommutative rings

Weaknesses

1. Stringent Conditions: Required conditions are quite special, limiting the universality of results
2. Computational Complexity: Practical verification of conditions requires complex number-theoretic calculations
3. Limited Applications: Primarily theoretical results with unclear direct application scenarios

Impact

1. Theoretical Contribution: Provides new perspectives on Waring-type problems in noncommutative algebra
2. Methodological Value: Proof techniques may apply to other related problems
3. Subsequent Research: Provides foundation and direction for further investigation

Applicable Scenarios

1. Algebraic Number Theory Research: Provides tools for related number-theoretic problems
2. Noncommutative Algebra: Enriches the content of noncommutative ring theory
3. Pure Mathematics: Theoretical mathematics research

References

The paper cites 11 important references, including:

1. Dickson's classical work on cyclotomy
2. Griffin-Krusemeyer's pioneering research
3. Ralph Hull's foundational theory on maximal orders in cyclic division algebras
4. Previous work by the authors in related areas

Overall Assessment: This is a high-quality number theory paper with significant technical innovation and important theoretical generalization. Although the conditions for the results are somewhat restrictive, it makes important contributions to the development of the related field. The paper's proof techniques are sophisticated and establish a foundation for subsequent research.