Sharp \(L^p-L^q\) estimates for Bochner--Riesz operators of negative index in \({\mathbb R}^n\), \(n\geqslant 3\) (Q705332): Difference between revisions

The authors consider that the Bochner-Riesz operator \(T^{-\alpha}\) on \(\mathbb{R}^n\), \(n \geq 3\), of order \(\alpha\) is a multiplier operator defined by \[ \widehat{T^{-\alpha}f}(\xi) = \frac{(1-| \xi| ^2)_{+}^{-\alpha}}{\Gamma(-\alpha + 1)} \, \hat{f}(\xi), \;\;\xi \in \mathbb{R}^n, \;\;\alpha > 0. \] Here \ \(\widehat{}\) \ denotes the Fourier transform, \(\Gamma\) is the gamma function and \(r_{+} = r\) if \(r \geq 0\) and \(r_{+} = 0\) if \(r < 0\). Let \[ \Delta_{\alpha} = \left\{ (\frac{1}{p}, \frac{1}{q}) \in [0,1] \times [0,1] \, : \, \frac{1}{p}- \frac{1}{q} \geq \frac{2\alpha}{n+1}, \, \frac{1}{p} > \frac{n-1}{2n} + \frac{\alpha}{n}, \, \frac{1}{q} < \frac{n+1}{2n} - \frac{\alpha}{n}\right\}. \] It is well known that being \((\frac{1}{p}, \frac{1}{q}) \in \Delta_{\alpha}\) is the necessary condition for \(L^p\)-\(L^q\) boundedness of \(T^{-\alpha}\). In the plane, Bak proved that the necessary condition is also a sufficient one for \(0 < \alpha \leq 3/2\) (see, [1]). In higher dimensions \(n \geq 3\), this is known for \(1/2 \leq \alpha < (n+1)/2\) (see [1,2,3]). In this paper, the authors extended this result to \(\frac{n^{2}-n-2}{2(n^{2}+ n -2)} < \alpha < 1/2\). To prove this, they first consider the Bochner-Riesz operator of negative index associated with a smooth hypersurface \((\xi', \psi(\xi'))\), \(\xi' \in \mathbb{R}^{n-1}\), where \(\psi\) is a smooth function on \([-1,1]^{n-1}\). More precisely, they define a multiplier operator defined by \[ \widehat{S^{-\alpha}f}(\xi) = \frac{(\xi_n -\psi(\xi'))_{+}^{-\alpha}}{\Gamma(-\alpha + 1)} \, \chi_{0}(\xi) \, \widehat{f}(\xi), \;\;(\xi',\xi_n) \in \mathbb{R}^{n-1} \times \mathbb{R}, \;n \geq 2, \] where \(\chi_{0}\) is a smooth function supported in a small neighborhood of the origin in \(\mathbb{R}^n\). By imposing an elliptic condition on \(\psi\), they show that the results for \(S^{-\alpha}\) imply the corresponding result for \(T^{-\alpha}\). References: [1] [\textit{J.-G. Bak}, Proc. Am. Math. 125, No. 7, 1977--1986 (1997; Zbl 0871.42012)] [2] [\textit{J.-G. Bak, D. McMichael} and \textit{D. Oberlin}, J. Aust. Math. Soc., Ser. A 58, No. 2, 154--166 (1995; Zbl 0829.46016)] [3] [\textit{S. Gutierrez}, Proc. Am. Math. 128, No. 2, 495--501 (2000; Zbl 0932.42016)].