← All problems

Goldberg's counterexample-existence question for the endpoint weighted Fourier estimate

Let n≥2n\ge 2 and pc=2(n+1)n+3p_c=\frac{2(n+1)}{n+3}. Does there exist a constant Cn<∞C_n<\infty such that every f∈Lpc(Rn)f\in L^{p_c}(\mathbb{R}^n) whose continuous Tomas--Stein spherical Fourier restriction vanishes, f^∣Sn−1=0\widehat f\vert_{\mathbb{S}^{n-1}}=0, satisfies

∥∣1−∣ξ∣2∣−1/2f^(ξ)∥L2(Rn)≤Cn∥f∥Lpc(Rn)?\left\|\left|1-|\xi|^2\right|^{-1/2}\widehat f(\xi)\right\|_{L^2(\mathbb{R}^n)}\le C_n\|f\|_{L^{p_c}(\mathbb{R}^n)}?
References

Source: arXiv

Progress summary

Refreshed
Open · solution reported

An unrefereed preprint claims to rule out the endpoint counterexamples Goldberg asked about, but the claim has not been independently verified.

Goldberg’s Problem 4 asks whether, at the critical endpoint, vanishing of the Fourier transform on the unit sphere guarantees the corresponding weighted square-integrability estimate. The question was recorded in 2017, with the endpoint left unresolved.

Known results

  • The sufficient estimate was known for 0≤α<120\leq\alpha<\frac12 and 12<α<32\frac12<\alpha<\frac32; the critical value α=12\alpha=\frac12 was excluded.

Endpoint preprint

Wenjuan Li, Manli Song, Hongqiang Wang, and Huiju Wang claim that at α=12\alpha=\frac12, the weighted estimate holds for functions whose spherical Fourier restriction vanishes continuously on the unit sphere. This gives a negative answer to Goldberg’s counterexample-existence question, but the evidence is an unrefereed preprint.

Current status (as of October 2026): Goldberg’s endpoint question is claimed solved negatively by an unrefereed preprint, while independent verification is absent.

Sources

Solutions 0

No solutions have been posted yet.