Answers are grounded in this paper and cite their sources.
Salem's equivalence states that the Riemann hypothesis is true if and only if, for every 12<k<1, every bounded measurable function f satisfying ∫0∞uk−1f(u)eut+1 du=0,t>0, vanishes almost everywhere on R+. Recent approaches to this equivalence have studied inversion formulae for the associated integral transform through Lambert-type and Widder–Lambert-type methods. In this paper, we develop a Fourier inversion approach to the Salem transform F(t)=∫0∞uk−1f(u)eut+1 du,t>0, 12<k<1. Under the Fourier invertibility assumptions Hf(r)=e−krf(e−r)∈L1(R),Hf^∈L1(R), we prove the inversion formula f(x)=−x−k2π∫−∞∞∫0∞F′(t)t−iξ dtΓ(1−iξ)η(−iξ)x−iξ dξ,x>0, valid almost everywhere, and pointwise at the corresponding continuity points of Hf. Consequently, if F(t)=0 for all t>0, then f = 0 almost everywhere in this Fourier invertible class. Thus the present result rules out an explicit family of bounded measurable functions as possible counterexamples to Salem's integral equation.
No takes yet. Share an insight, caveat, or question.
Irshad Ayoob (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: