Strong Invariance Principle for Random Dirichlet Series
summary
The gist
It is shown that the random prime Dirichlet series X(s) = P p∈P ξp ps+1/2 can be almost surely approximated by a Brownian motion…
In short
The paper investigates how random prime Dirichlet series, $X(s)$, can be approximated by a Brownian motion under certain conditions involving independent random variables. It establishes that this approximation leads to important limit laws like the functional law of the iterated logarithm. Furthermore, it shows that a related series $Y(s)$ shares these same properties, confirming shared asymptotic behavior.
Key concepts
- Random Prime Dirichlet Series $X(s)$
- $X(s) = \sum_{p \in \mathcal{P}} \frac{\xi_p}{p^{s+1/2}}$ is a series where the coefficients $\xi_p$ are random variables associated with prime numbers. The study aims to understand the statistical behavior of this series as $s$ approaches 0, specifically its connection to Brownian motion.
- Strong Invariance Principle
- This principle demonstrates that a complex random process, like $X(s)$, can be strongly approximated by a simpler Gaussian process, such as a standard Brownian motion. This approximation holds almost surely under specific conditions on the random variables, allowing researchers to use tools from Brownian motion theory to analyze the series' asymptotic behavior.
- Functional Law of the Iterated Logarithm (LIL)
- This is a powerful limit law that describes the precise rate at which a stochastic process fluctuates over time. The paper derives this law for $X(s)$ and $Y(s)$, showing how the fluctuations of these random series behave as they approach their limits, providing detailed information about their long-term statistical trends.
Terminology used across episodes
This episode discusses
- Strong Invariance Principle for Random Dirichlet Series · Paper Radio
- Limit theorems for random Dirichlet series with summation over primes, with an application to Rademacher random multiplicative functions
- On the maxima of Littlewood polynomials on [-1,1]
The paper
Strong Invariance Principle for Random Dirichlet Series · Read on arXiv
Li-Xin ZHANG, Zhejiang Gongshang University
Zhejiang Gongshang University
Transcript
Introduction to the show: ident: Mathematics. The day in mathematics: new theorems, new proofs, and what they open.
Ada: I'm Ada.
Emmy: And I'm Emmy, here with Kolya.
Ada: We're looking at the paper: "Strong Invariance Principle for Random Dirichlet Series".
Emmy: Here's the gist of the paper. It is shown that the random prime Dirichlet series X(s) = P p∈P ξp ps+1/2 can be almost surely approximated by a Brownian motion, where there are independent and identically distributed random variables with zero mean and finite variance, and P is the set of prime numbers The paper establishes that the random Dirichlet series X(s) can be approximated by a standard Brownian motion under certain conditions The main theorem states that there exists a standard Brownian motion B(t) such that X(s) − σB(log(1/s)) = o p log log log(1/s) a.s.
Ada: as s → 0 Furthermore, if E
ξ squared (log ξ) 1+ϵ: 0, then X(s) − σB(log(1/s)) = O p log log(1/s) a.s.
Emmy: as s → 0 Approximation for Y (s) The paper also shows that the random Dirichlet series Y (s) = X∞ k=2 (log k) α ks+1/2 ξk with α = −1/2 admits the same approximation as X(s), and a direct relationship between X(s) and Y (s) with α = −1/2 is established For Y (s) with α > −1/2, it is shown that Y (s) can be approximated by a Laplace transform of a power-weighted Brownian stochastic integral Functional Limit Theorems The strong invariance principle allows for the derivation of several important limit laws for X(s) and Y(s) Specifically, the functional law of the iterated logarithm, limit laws of the iterated logarithm, and Chung’s law of the iterated logarithm for X(s) are obtained For Y (s) with α = −1/2, similar properties hold When α > −1/2, a functional CLT and functional LIL are obtained for X(st) Special Cases and Proof Techniques The proof of the strong invariance principle relies on a two-stage approximation approach: first, replacing random variables with normal random variables to form a Gaussian process, and then approximating this Gaussian process by a Brownian motion or its Laplace transform The paper considers the special case where the random variables are normal and proves that Z(s) = X p∈P ηp ps+1/2 can be approximated by a Brownian motion For the general case, it is shown that a sequence of independent N(0, 1) random variables can be constructed such that X∞ k=1 ξk − σηk dk(s) = o p log(1/s) log log log(1/s) a.s.
Ada: Relationship between Series A direct relationship is established between X(s) and Y (s), where Y1(s) = X∞ k=2 ξk p s+1/2 k, and Y2(s) = X∞ k=2 ξk (k log k) s+1/2 This relationship is crucial for showing that both series share the same functional limit laws The proof of Theorem 4.1 shows that Yi(s) − Y (s) = o p log(1/s) log log log(1/s) a.s.
Paper summary: Ada: Okay, so to recap, this paper "Strong Invariance Principle for Random Dirichlet Series" is showing that the random prime Dirichlet series X(s) = sum p in P xi p over p s+one/two can be approximated by a standard Brownian motion almost surely <ref:2610.18353#pg1,that the random prime Dirichlet series $X(s>.
Emmy: The thesis is establishing this strong approximation under conditions where the xi p are independent and identically distributed random variables with zero mean and finite variance, and P is just the set of prime numbers.
Ada: As a consequence of this approximation, they obtain several important limit laws: the functional law of the iterated logarithm, limit laws of the iterated logarithm, and Chung’s law of the iterated logarithm for X(s).
Emmy: They also look at a modified version called Y(s), which is defined as Y(s) = sum k=two infinity(k) alpha over k s+one/two xi k, and they show it admits the same approximation as X(s) when alpha is set to negative one-half <ref:2610.18353#pg1,admits the same approximation as $X(s>.
Ada: They also establish a direct relationship between X(s) and Y(s) specifically when alpha equals negative one-half. Furthermore, they obtain the strong invariance principle for Y(s) for cases where alpha is greater than negative one-half.
Emmy: This work connects the behavior of these two different types of random Dirichlet series, showing they share these important limit laws and that both can be strongly approximated by Brownian motion or its Laplace transform.
Kolya: So, focusing on the structure, what's the main point here? It sounds like they are saying that despite the randomness coming from primes and other factors in Y(s), the underlying scaling behavior is controlled by a standard Brownian motion.
Ada: Exactly. The underlying structure is robust enough to be captured by a standard Brownian motion, which is what allows them to derive these functional limit theorems and limit laws for X(s).
Emmy: It suggests that the randomness doesn't completely destroy the predictable, continuous path behavior of the process; it just modulates it in a way that fits this known stochastic model.
Kolya: And for my part, I’m thinking about the tails. When they prove these laws hold, what does that say about which tail governs this approximation? Is it governed by something heavier than a simple Gaussian distribution?
Ada: The results are derived from the limit laws of iterated logarithm and Chung’s law of the iterated logarithm for Brownian motion itself, so the structure is essentially governed by those known properties.
Emmy: It connects probabilistic number theory to continuous analysis in a very concrete way, which is what makes this paper interesting from an algebraic geometry viewpoint.
Kolya: Does this mean that if we applied this to something else, say a Dirichlet series with different coefficients, we could expect similar results? What are the boundaries of the approximation?
Ada: The paper states that for X(s), they can construct a sequence of independent normal random variables to get an approximation. For the general case, they show you can build such a sequence from independent N(zero one) random variables such that X(s) - sigma B((one/s)) = o p (one/s) (one/s) almost surely <ref:2610.18353#pg3,log(1/s)) = o p \log(1/s) \log\log\log(1>.
Emmy: That error term is quite specific. It tells you exactly how fast the error decays as s approaches zero, which is what lets them prove the strong approximation holds in a rigorous way.
Kolya: So, it's not just that it *eventually* gets close; it's that there’s a quantifiable rate of convergence based on those log terms.
Ada: That quantifiable rate is what allows them to transition from the approximation itself to the actual functional laws they are interested in proving.
Conclusion: Ada: So, looking at the whole paper "Strong Invariance Principle for Random Dirichlet Series", it establishes that both X(s) and Y(s) can be approximated by Brownian motion or its Laplace transform.
Emmy: This means we have established the strong invariance principle for these series. It confirms they can be modeled by this continuous stochastic process, which implies the functional central limit theorems and law of the iterated logarithm for both X(s) and Y(s).
Kolya: The implication is that whatever randomness is happening in these sums, its large-scale behavior settles down to a predictable path defined by Brownian motion.
Ada: It confirms prior conjectures about the shared limit laws, suggesting that the asymptotic behavior of geometrically weighted random series is the same as that of a random Dirichlet series with alpha equal to zero.
Emmy: And it also provides a relationship between X(s) and Y(s) when alpha is negative one-half, which supports their shared limit laws by showing how they are mathematically linked.
Kolya: So
More episodes
- 2509.05860-Concentration Inequalities for Branching Random Walks
- 2509.19309-Small Prime Primitive Roots in Arithmetic Progressions
- 2509.21830-Noncollapsing for Curvature Flows with Inhomogeneous Speeds
- 2510.08346-Separability and Submetrizability in Locally Convex Spaces
- 2510.27290-Borel Combinatorics of Schreier Graphs of Z-actions
- 2511.13500-Sampling Density for Gabor Phase Retrieval
- 2511.19161-Linear dynamics of random products of weighted shifts
- 2511.21548-Metastability of diffusion processes in narrow tubes
- 2512.01536-Abelian approximations for bicrossed products
- 2601.02524-Lagrangian slice disks with symplectomorphic exteriors