One and two weight norm inequalities for Riesz potentials (Q2509825)

The question of sharp one and two weight norm inequalities for Calderón-Zygmund operators has been intensely studied in recent years. The problem of sharp constants in the one weight case, also known as the \(A_2\) conjecture, was solved in full generality by \textit{T. P. Hytönen} [Ann. Math. (2) 175, No. 3, 1473--1506 (2012; Zbl 1250.42036)]. Subsequently, \textit{T. P. Hytönen} and \textit{C. Pérez} [Anal. PDE 6, No. 4, 777--818 (2013; Zbl 1283.42032)] improved on this result by replacing a part of the \(A_p\) bounds by weaker \(A_\infty\) bounds in terms of Wilson's \(A_\infty\) constant \[ [w]_{A_\infty^{'}}=\sup_Q \frac{1}{w(Q)} \int_Q M(\chi_Q w)(x)\,dx<\infty. \] This constant can be substantially smaller than the \(A_\infty\) constant introduced by Hruščev. In this paper, the authors study analogous inequalities for the Riesz potentials \[ I_\alpha f(x)= \int_{\mathbb{R}^n} \frac{f(y)}{|x-y|^{n-\alpha}}, \quad 0<\alpha<n. \] Given the natural scaling of \(I_\alpha\), one considers the class of weights \(A_{p,q}\), defined such that \([w]_{A_{p,q}}=[u]_{A_{s(p)}}^{1/q}\) with \(u=w^q\) and \(s(p)=p\left(\frac{n-\alpha}{n-\alpha p}\right)\). The corresponding result of the one weight problem was already established by \textit{M. T. Lacey} et al. [J. Funct. Anal. 259, No. 5, 1073--1097 (2010; Zbl 1196.42014)]. The authors now give an improved result in parallel to the \(A_p\)-\(A_\infty\)-type estimate of Hytönen and Pérez. That is, for \(0<\alpha<n\), \(1<p<\frac{n}{\alpha}\) and \(\frac{1}{p}-\frac{1}{q}=\frac{\alpha}{n}\), they show that \[ \|I_\alpha(\cdot \sigma)\|_{L^p(\sigma) \to L^{q,\infty}(u)} \lesssim [u]_{A_{s(p)}}^{\frac{1}{q}} [u]_{A_\infty^{'}}^{\frac{1}{p'}}, \] where \(w \in A_{p,q}\), \(u=w^q\), \(\sigma=w^{-p'}\), together with a strong type inequality. They prove this result as a special case of a two weight result with a weakened coupling between the weights \(u\) and \(\sigma\), which parallels the result of \textit{T. P. Hytönen} and \textit{M. T. Lacey} [Indiana Univ. Math. J. 61, No. 6, 2041--2052 (2012; Zbl 1290.42037)] for Calderón-Zygmund operators. As for Calderón-Zygmund operators, the proofs rely on a corona decomposition, but adapted to the case of fractional integral operators. In a second part, the authors consider with similar methods the question of log bump conditions. They establish new weak and strong type inequalities in the two weight case under testing conditions involving Orlicz spaces, where the defining class of Young functions is of the form \(\Phi(t)=t^q\log(e+t)^{q-1+\delta}\), \(\delta>0\).