Some ergodic theorems over squarefree numbers and squarefull numbers

Basic Information

• Paper ID: 2405.18157
• Title: Some ergodic theorems over squarefree numbers and squarefull numbers
• Authors: Huixi Li, Biao Wang, Chunlin Wang, Shaoyun Yi
• Classification: math.NT (Number Theory), math.DS (Dynamical Systems)
• Publication Date: October 13, 2025
• Paper Link: https://arxiv.org/abs/2405.18157

Abstract

In 2022, Bergelson and Richter provided a new dynamical generalization of the Prime Number Theorem by establishing ergodic theorems along the number of prime factors of integers. They also proved that this generalization holds when integers are restricted to squarefree numbers. This paper introduces the concept of invariant averages of arithmetic functions under multiplication. Using properties of these invariant averages, the authors derive several ergodic theorems concerning squarefree numbers and squarefull numbers. These theorems are intimately connected with the Erdős-Kac theorem, the Bergelson-Richter theorem, and the Loyd theorem.

Research Background and Motivation

Problem Background

1. Dynamical generalization of the Prime Number Theorem: Bergelson and Richter established in 2022 a dynamical generalization of the Prime Number Theorem through ergodic theorems in the form: for any uniquely ergodic topological dynamical system $(X,\mu,T)$, $\lim_{N\to\infty} \frac{1}{N}\sum_{n=1}^N f(T^{\Omega(n)}x) = \int_X f d\mu$
2. Restriction to squarefree numbers: They further proved that similar results hold when the sequence is restricted to squarefree numbers: $\lim_{N\to\infty} \frac{1}{N}\sum_{\substack{1\leq n\leq N\\n \text{ squarefree}}} f(T^{\Omega(n)}x) = \frac{6}{\pi^2}\int_X f d\mu$

Research Motivation

1. Theoretical refinement: Although the Bergelson-Richter work pioneered the establishment of a dynamical form of the Prime Number Theorem, there is a lack of systematic research on ergodic theorems over special number sets (such as squarefree numbers and squarefull numbers).
2. Methodological unification: Various existing ergodic theorems (such as the Loyd theorem and generalizations by Wang et al.) lack a unified theoretical framework; a more general approach is needed to handle such problems.
3. Number-theoretic applications: Squarefree numbers and squarefull numbers occupy important positions in number theory, and their distribution properties are closely related to core concepts such as the Prime Number Theorem and the Riemann zeta function.

Core Contributions

1. Introduction of invariant average concept: Proposes the concept of invariant averages of arithmetic functions under multiplication, providing a unified framework for studying ergodic theorems over special number sets.
2. Establishment of main theorems: Proves the main theorem for squarefree numbers (Theorem 1.1) and the main theorem for k-full numbers (Theorem 1.4).
3. Generalization of classical results: Extends the Erdős-Kac theorem, the Bergelson-Richter theorem, the Loyd theorem, and others to squarefree numbers and squarefull numbers.
4. Provision of applications: Provides analogues of Richter's generalization of the Prime Number Theorem over special number sets (Theorem 1.7).

Detailed Methodology

Core Concept Definitions

Definition of invariant average: For a bounded arithmetic function $a: \mathbb{N} \to \mathbb{C}$, it is said to possess an invariant average under multiplication if the average $\lim_{N\to\infty} \frac{1}{N}\sum_{n=1}^N a(n)$ exists and satisfies $\lim_{N\to\infty} \frac{1}{N}\sum_{n=1}^N a(mn) = \lim_{N\to\infty} \frac{1}{N}\sum_{n=1}^N a(n)$ for all $m \in \mathbb{N}$.

k-th power invariant average: Similarly, the k-th power invariant average is defined as $\lim_{N\to\infty} \frac{1}{N}\sum_{n=1}^N a(n^k m) = \lim_{N\to\infty} \frac{1}{N}\sum_{n=1}^N a(n^k)$

Main Technical Tools

1. Treatment of squarefree numbers (Proposition 3.2)

Using properties of the Möbius function, for a bounded function $a$ and $1 \leq D \leq \sqrt{N}$: $\frac{1}{N}\sum_{n=1}^N \mu^2(n)w_S(n)a(n) = \sum_{d=1}^D \frac{\mu(d)w_S(d)}{d^2} \mathbb{E}_{n\in[N/d^2]} w_S(n)a(d^2n) + O\left(\frac{1}{D}\right) + O\left(\frac{1}{\sqrt{N}}\right)$

where $w_S(n)$ is the indicator function for integers not divisible by primes in the set $S$.

2. Treatment of k-full numbers (Proposition 5.1)

Using the unique representation of k-full numbers $n = m^k n_1^{k+1}\cdots n_{k-1}^{2k-1}$ (where $n_i$ are pairwise coprime and squarefree), a similar decomposition formula is obtained.

Proof Strategy

1. Establishment of decomposition formulas: Through Möbius inversion and structural decomposition of k-full numbers, sums over special number sets are transformed into standard sums.
2. Application of invariance: Using properties of invariant averages, the existence and values of limits are established.
3. Error estimation: Through careful error analysis, various remainder terms are controlled so that they vanish in the limit.

Main Results

Theorem 1.1 (Ergodic theorem over squarefree numbers)

Let $a: \mathbb{N} \to \mathbb{C}$ be a bounded arithmetic function with average $A$. Let $S$ be a finite set of primes. If the average of $a(n)$ is invariant under multiplication, then $\lim_{N\to\infty} \frac{1}{N}\sum_{\substack{1\leq n\leq N\\n \text{ squarefree}\\p\nmid n, \forall p\in S}} a(n) = \frac{\alpha(S)}{\zeta(2)}A$

where $\alpha(S) = \prod_{p\in S} \frac{p}{p+1}$. In particular, when $S = \emptyset$: $\lim_{N\to\infty} \frac{1}{N}\sum_{\substack{1\leq n\leq N\\n \text{ squarefree}}} a(n) = \frac{6}{\pi^2}A$

Theorem 1.4 (Ergodic theorem over k-full numbers)

Let $k \geq 2$. If $a: \mathbb{N} \to \mathbb{C}$ is a bounded arithmetic function with k-th power invariant average, then $\lim_{N\to\infty} \mathbb{E}_{n\in[N], n \text{ is } k\text{-full}} a(n) = \lim_{N\to\infty} \mathbb{E}_{n\in[N]} a(n^k)$

Application Examples

1. Generalization of the Erdős-Kac theorem

For squarefree numbers: $\lim_{N\to\infty} \frac{1}{N}\sum_{\substack{1\leq n\leq N\\n \text{ squarefree}}} F\left(\frac{\Omega(\phi^k(n)) - a_k(\log\log N)^{k+1}}{b_k(\log\log N)^{k+1/2}}\right)f(T^{\Omega(n)}x) = \frac{6}{\pi^2}\left(\frac{1}{\sqrt{2\pi}}\int_{-\infty}^{\infty} F(t)e^{-t^2/2}dt\right)\left(\int_X f d\mu\right)$

2. Results over squarefull numbers

For squarefull numbers, when $(X,\mu,T)$ is a completely uniquely ergodic system: $\lim_{N\to\infty} \mathbb{E}_{\substack{n\in[N]\\n \text{ is } k\text{-full}}} F\left(\frac{\Omega(n) - k\log\log N}{k\sqrt{\log\log N}}\right)f(T^{\Omega(n)}x) = \left(\frac{1}{\sqrt{2\pi}}\int_{-\infty}^{\infty} F(t)e^{-t^2/2}dt\right)\left(\int_X f d\mu\right)$

Technical Innovations

1. Introduction of the invariant average concept

This is the core innovation of the paper. By introducing the concept of invariant averages under multiplication, the authors provide a unified framework for handling various ergodic theorems. The key aspects of this concept are:

• Unification: Many known arithmetic functions satisfy this property
• Verifiability: For concrete functions, invariance can be verified through existing number-theoretic techniques
• Transitivity: Invariance is preserved under certain transformations

2. Improvements in technical methods

• Refined error estimates: When treating squarefree numbers and k-full numbers, the authors provide precise error bounds
• Unified proof strategy: Through Möbius inversion and properties of number-theoretic functions, problems are reduced to standard cases

3. Breadth of results

The paper not only addresses individual ergodic theorems but systematically generalizes multiple important theorems, including:

• The Bergelson-Richter theorem
• The Erdős-Kac theorem
• The Loyd theorem
• Richter's generalization of the Prime Number Theorem

Related Work

1. Bergelson-Richter theorem (2022)

Established a dynamical generalization of the Prime Number Theorem, serving as the direct starting point for this paper.

2. Erdős-Kac theorem (1940)

Proved that $\Omega(n)$ follows a normal distribution, a foundational work in probabilistic number theory.

3. Loyd theorem (2023)

Proved that the sequences in the Erdős-Kac theorem and the Bergelson-Richter theorem are asymptotically uncorrelated.

4. Related generalizations

• Wang (2022): Generalizations on additive arithmetic semigroups
• Wang et al. (2025): Dynamical version of the Erdős-Pomerance theorem

Conclusions and Discussion

Main Conclusions

1. Theoretical contribution: Establishes a unified theoretical framework for handling ergodic theorems over special number sets
2. Technical progress: Develops new number-theoretic techniques for handling sums restricted to special sets
3. Applied value: Provides new tools for cross-disciplinary research between probabilistic number theory and ergodic theory

Limitations

1. Technical constraints: The methods are primarily applicable to number sets with good multiplicative structure
2. Condition requirements: Functions must satisfy the invariant average condition, which may be difficult to verify in some cases
3. Scope of generalization: For more general number sets (such as r-free numbers, $r \geq 3$), further research is needed for extensions of the results

Future Directions

1. Infinite prime sets: Consider cases where the set $S$ is infinite
2. Finite field analogues: Study similar results over finite fields
3. More general number sets: Extend to r-free numbers and other number sets with special structure

In-Depth Evaluation

Strengths

1. Theoretical innovation: The introduction of the invariant average concept is a genuine innovation, providing new research tools for the field
2. Technical sophistication: The proof techniques are skillful, error estimates are precise, and reflect high-level number-theoretic expertise
3. Rich results: Not only provides main theorems but also offers numerous applications and generalizations
4. Clear exposition: The paper is well-structured, technical details are clear, and it is easy to understand and verify

Weaknesses

1. Limited applications: Verification of the invariant average condition may be difficult in some cases
2. Computational complexity: Practical computation of some results may involve complex number-theoretic functions
3. Generalization potential: While results are abundant, directions for further generalization are not sufficiently clear

Impact

1. Academic value: Opens new directions for cross-disciplinary research between ergodic theory and number theory
2. Technical impact: The provided technical methods may be adopted and developed by other researchers
3. Theoretical significance: Deepens understanding of dynamical generalizations of the Prime Number Theorem

Applicable Scenarios

1. Theoretical research: Applicable to studying distribution properties over special number sets
2. Probabilistic number theory: Provides new proof methods for limit theorems in probabilistic number theory
3. Ergodic theory: Provides new number-theoretic backgrounds for applications of ergodic theory

References

The paper cites 35 relevant references, covering important works in ergodic theory, number theory, probabilistic number theory, and other fields, reflecting the depth and breadth of the research. Main references include:

• Pioneering works by Bergelson and Richter
• The Erdős-Kac theorem and its generalizations
• Classical results on the distribution of squarefree numbers and squarefull numbers
• Recent advances at the intersection of dynamical systems and number theory