← All problems

Hagelstein–Parissis sharp Solyanik-exponent question

For n≥1n\ge 1, determine the optimal exponent γn\gamma_n governing the sharp Tauberian constants of the discrete and ergodic uncentered Hardy--Littlewood maximal operators as α→1−\alpha\to1^{-}; namely, determine whether γn\gamma_n is characterized by C~HL(α)−1≍n(1/α−1)γn\widetilde C_{\mathrm{HL}}(\alpha)-1\asymp_n(1/\alpha-1)^{\gamma_n} and, for the worst-case ergodic constant, by CHL∗(α)−1≍n(1/α−1)γnC^{*}_{\mathrm{HL}}(\alpha)-1\asymp_n(1/\alpha-1)^{\gamma_n}. Hagelstein and Parissis proved the upper bound CHL∗(α)−1≲n(1/α−1)1/(n(n+1))C^{*}_{\mathrm{HL}}(\alpha)-1\lesssim_n(1/\alpha-1)^{1/(n(n+1))} and asked whether the exponent 1/(n(n+1))1/(n(n+1)) is sharp. Huang claims that the sharp discrete exponent, and the sharp exponent for ergodic actions admitting arbitrarily large Rokhlin towers, is γn=1/n\gamma_n=1/n.

Equivalent formulations
  1. Is the exponent 1/(n(n+1))1/(n(n+1)) in the estimate CHL∗(α)−1≲n(1/α−1)1/(n(n+1))C^{*}_{\mathrm{HL}}(\alpha)-1\lesssim_n(1/\alpha-1)^{1/(n(n+1))} sharp as α→1−\alpha\to1^{-}?

  2. Determine the sharp asymptotic behavior of the discrete Tauberian constant C~HL(α)\widetilde C_{\mathrm{HL}}(\alpha) as α→1−\alpha\to1^{-}; Huang claims C~HL(α)−1≍n(1/α−1)1/n\widetilde C_{\mathrm{HL}}(\alpha)-1\asymp_n(1/\alpha-1)^{1/n}.

    source: Huang, Kesen

References

Progress summary

Refreshed
Open · progress reported

A new preprint claims the lattice and ergodic versions have a sharp exponent, while the corresponding Euclidean question has a different answer.

Hagelstein and Parissis posed the sharpness question as Problem 7.27.2. A new preprint claims that the discrete and worst-case ergodic exponent is 1/n1/n, rather than the earlier weaker exponent.

Known results

  • Hagelstein–Parissis: the earlier ergodic estimate had exponent 1/(n(n+1))1/(n(n+1)).
  • Mukherjee: the Euclidean estimate is CHL(α)−1≍n(1−α)2/(n+1)C_{\mathrm{HL}}(\alpha)-1\asymp_n(1-\alpha)^{2/(n+1)} for n≥2n\ge 2.

October 2026 sharpness claim

Kesen Huang’s preprint claims C~HL(α)−1≍n(1−α)1/n\widetilde C_{\mathrm{HL}}(\alpha)-1\asymp_n(1-\alpha)^{1/n} for α\alpha close to 11, with transference to ergodic actions admitting arbitrarily large Rokhlin towers. This claims sharpness for the discrete and worst-case ergodic problems, not for the Euclidean constants; independent verification was not found.

Current status (as of November 2026): The discrete and worst-case ergodic exponent is claimed to be sharp at 1/n1/n, while the Euclidean exponent is separately known as 2/(n+1)2/(n+1); the Hagelstein–Parissis question is not independently verified as settled.

Sources

Solutions 0

No solutions have been posted yet.