On the Fourier transform of the characteristic functions of domains with \(C^1\) boundary (Q372460)

Based on excerpts from the author's introduction: We denote the boundary of a domain (an open connected set) \(D\subseteq\mathbb{R}^{n}\) by \(\partial D.\) Saying that the boundary of \(D\) is \(C^{1}\) or \(C^{2},\) we mean that each of its points has a neighbourhood in which the boundary \( \partial D\) is the graph of certain function (real valued) of class \(C^{1}\) or \(C^{2}\). For each domain \(D\subseteq\mathbb{R}^{n}\) with\ \(C^{1}\) boundary, let \(v_{D}\left( x\right) \) be the outer unit normal vector to \(\partial D\) at a point \(x\in \partial D.\) The corresponding map \(v_{D}:\partial D\rightarrow S^{n-1}\) is called normal map. By \(\omega \left( v_{D},\delta \right) \) we denote the modulus \ of continuity of \( v_{D}:\) \[ \omega \left( v_{D},\delta \right) =\sup_{x,y\in \partial D;\left| x-y\right| \leq \delta }\left| v_{D}\left( x\right) -v_{D}\left( y\right) \right| ,\;\delta \geq 0, \] where \(\left| u\right| \) is the lenght of \(u\in\mathbb{R}^{n}\). Let \(\omega \left( \delta \right) \) be an arbitrary nondecreasing continuous function on \([0,\infty )\) with \(\omega \left( 0\right) =0.\) In case when \(\omega \left( v_{D},\delta \right) =O\left( \omega \left( \delta \right) \right) ,\delta \rightarrow +0,\) we say that the boundary\ \(\partial D\) is \(C^{1,\omega }\). Let \(D\subseteq\mathbb{R}^{n}\) be a bounded domain (an open connected set), \(n\geq 2.\) Consider its characteristic function \(1_{D,\text{ }}\)and the Fourier transform \(\overset{ \wedge }{1_{D}}\) of this function. We define the spaces \(A_{p}\left(\mathbb{R}^{n}\right)\), \(1\leq p\leq \infty,\) of tempered distributions \(f\) on \(\mathbb{R}^{n}\) such that the Fourier transform \(\overset{\wedge }{f}\) belongs to \( L^{p}\left(\mathbb{R}^{n}\right)\). The norm on \(A_{p}\left(\mathbb{R}^{n}\right) \) is defined as \[ \left\| f\right\| _{A_{p}\left(\mathbb{R}^{n}\right) }=\left\| \overset{\wedge }{f}\right\| _{L^{p}\left(\mathbb{R}^{n}\right) }. \] In this paper in Section 1 the author gives a proof of the relation \(1_{D}\in A_{p}\left(\mathbb{R}^{n}\right) ,\) which holds for all \(p>\frac{2n}{n+1}\) whenever \(D\subseteq\mathbb{R}^{n}\) is a bounded domain \ with \(C^{1}\) boundary. In Section 2 the author gives the main result of this paper. Namely, it is shown that if \(\partial D\in C^{1,\omega }\) and \[ \int_{0}^{1}\frac{\delta ^{n\left( p-1\right) -1}}{\omega \left( \delta \right) ^{n-p}}d\delta =\infty , \] then \(1_{D}\notin A_{p}\left(\mathbb{R}^{n}\right) .\) In particular , if \(\partial D\in C^{1,\alpha },\) then \( 1_{D}\notin A_{p}\left(\mathbb{R}^{n}\right) \) for \[ p\leq 1+\frac{(n-1)\alpha }{n+\alpha }. \] In Section 3 the author proves that if \(D\subseteq\mathbb{R}^{2}\) is\(\;\)a domain such that \(\partial D\in C^{1,\omega }\) and \[ \int_{0}^{1}\frac{\delta ^{2p-3}}{\omega \left( \delta \right) ^{2-p} }d\delta =\infty , \] then \(1_{D}\notin A_{p}\left(\mathbb{R}^{2}\right)\). In particular, if \(\partial D\in C^{1,\alpha },\) then \( 1_{D}\notin A_{p}\left(\mathbb{R}^{2}\right) \) for \(p\leq 1+\frac{\alpha }{2+\alpha }\). The author also shows that this result is sharp, namely, for each class \(C^{1,\omega }\) (under a certain simple condition imposed on \(\omega ),\) there exists a bounded domain \(D\subseteq\mathbb{R}^{2}\) such that \(\partial D\in C^{1,\omega }\) and, for all \(p>1\) satisfying \[ \int_{0}^{1}\frac{\delta ^{2p-3}}{\omega \left( \delta \right) ^{2-p} }d\delta <\infty , \] then \(1_{D}\in A_{p}\left(\mathbb{R}^{2}\right)\). In particular, if \(0<\alpha <1,\) then there exists a planar domain \(D\) with \(C^{1,\alpha }\) boundary such that \(1_{D}\in A_{p}\left(\mathbb{R}^{2}\right) \) for all \(p>1+\frac{\alpha }{2+\alpha }.\) It follows that there exists a planar domain \(D\) with \(C^{1}\) boundary such that \(1_{D}\in A_{p}\left(\mathbb{R}^{2}\right) \) for all \(p>1\).