On the supremum of random Dirichlet polynomials

Mikhail Lifshits; Michel Weber

Studia Mathematica (2007)

  • Volume: 182, Issue: 1, page 41-65
  • ISSN: 0039-3223

Abstract

top
We study the supremum of some random Dirichlet polynomials D N ( t ) = n = 2 N ε d n - σ - i t , where (εₙ) is a sequence of independent Rademacher random variables, the weights (dₙ) are multiplicative and 0 ≤ σ < 1/2. Particular attention is given to the polynomials n τ ε n - σ - i t , τ = 2 n N : P ( n ) p τ , P⁺(n) being the largest prime divisor of n. We obtain sharp upper and lower bounds for the supremum expectation that extend the optimal estimate of Halász-Queffélec, s u p t | n = 2 N ε n - σ - i t | ( N 1 - σ ) / ( l o g N ) . The proofs are entirely based on methods of stochastic processes, in particular the metric entropy method.

How to cite

top

Mikhail Lifshits, and Michel Weber. "On the supremum of random Dirichlet polynomials." Studia Mathematica 182.1 (2007): 41-65. <http://eudml.org/doc/284720>.

@article{MikhailLifshits2007,
abstract = {We study the supremum of some random Dirichlet polynomials $D_\{N\}(t) = ∑_\{n=2\}^\{N\} εₙdₙn^\{-σ-it\}$, where (εₙ) is a sequence of independent Rademacher random variables, the weights (dₙ) are multiplicative and 0 ≤ σ < 1/2. Particular attention is given to the polynomials $∑_\{n∈ _\{τ\}\} εₙn^\{-σ-it\}$, $_\{τ\} = \{2 ≤ n ≤ N : P⁺(n) ≤ p_\{τ\}\}$, P⁺(n) being the largest prime divisor of n. We obtain sharp upper and lower bounds for the supremum expectation that extend the optimal estimate of Halász-Queffélec, $ sup_\{t∈ ℝ\} |∑_\{n=2\}^\{N\} εₙn^\{-σ-it\}| ≈ (N^\{1-σ\})/(log N)$. The proofs are entirely based on methods of stochastic processes, in particular the metric entropy method.},
author = {Mikhail Lifshits, Michel Weber},
journal = {Studia Mathematica},
keywords = {Dirichlet polynomials; Rademacher random variables; metric entropy method},
language = {eng},
number = {1},
pages = {41-65},
title = {On the supremum of random Dirichlet polynomials},
url = {http://eudml.org/doc/284720},
volume = {182},
year = {2007},
}

TY - JOUR
AU - Mikhail Lifshits
AU - Michel Weber
TI - On the supremum of random Dirichlet polynomials
JO - Studia Mathematica
PY - 2007
VL - 182
IS - 1
SP - 41
EP - 65
AB - We study the supremum of some random Dirichlet polynomials $D_{N}(t) = ∑_{n=2}^{N} εₙdₙn^{-σ-it}$, where (εₙ) is a sequence of independent Rademacher random variables, the weights (dₙ) are multiplicative and 0 ≤ σ < 1/2. Particular attention is given to the polynomials $∑_{n∈ _{τ}} εₙn^{-σ-it}$, $_{τ} = {2 ≤ n ≤ N : P⁺(n) ≤ p_{τ}}$, P⁺(n) being the largest prime divisor of n. We obtain sharp upper and lower bounds for the supremum expectation that extend the optimal estimate of Halász-Queffélec, $ sup_{t∈ ℝ} |∑_{n=2}^{N} εₙn^{-σ-it}| ≈ (N^{1-σ})/(log N)$. The proofs are entirely based on methods of stochastic processes, in particular the metric entropy method.
LA - eng
KW - Dirichlet polynomials; Rademacher random variables; metric entropy method
UR - http://eudml.org/doc/284720
ER -

NotesEmbed ?

top

You must be logged in to post comments.