Unimodular Fourier multipliers on Wiener amalgam spaces (Q2252294): Difference between revisions

A Fourier multiplier \(\sigma \left( D\right) \) in \(\mathbb{R}^{n}\) is an operator whose action on a test function \(f\) is defined by \[ \sigma \left( D\right) f\left( x\right) =\frac{1}{\left( 2\pi \right) ^{n}} \int\limits_{\mathbb{R}^{n}}e^{i\xi .x}\sigma \left( \xi \right) \hat{f}\left( \xi \right) d\xi. \] One can rewrite this operator as a convolution operator \(\sigma \left( D\right) f\left( x\right) =\check{\sigma}\ast f\left( x\right) \), where \( \check{\sigma}\) is the (distributional) inverse Fourier transform. The function \(\sigma \) is called the symbol or multiplier. Unimodular Fourier multipliers are defined by symbols of the type \(\sigma \left( D\right) =e^{i\left| D\right| ^{\alpha }}.\) The fundamental problem in the study of Fourier multipliers is to relate the boundedness of \(\sigma \left( D\right) \) on certain spaces to that of the properties of the symbol \(\sigma .\) It is known that the operator \(\sigma \left( D\right) =e^{i\left| D\right| ^{\alpha }}\) is bounded on \(L^{p}\) if and only if \(p=2\) see [\textit{L. Hörmander}, Acta Math. 104, 93--140 (1960; Zbl 0093.11402)]. The goal of this paper is to study the boundedness of unimodular Fourier multipliers \(\sigma \left( D\right) =e^{i\left| D\right| ^{\alpha }}\) with \(\alpha >0\) on Wiener amalgam space \(W^{p,q}.\) The authors prove the following main theorems. Theorem. Let \(\alpha \geq 2\) and \(\mu \) be a real-valued homogeneous function on \(\mathbb{R}^{n}\) of degree \(\alpha \) which belongs to \(C^{\infty }\left(\mathbb{R}^{n}\backslash \left\{ 0\right\} \right)\). Let \(1\leq p,q\leq \infty \) and \( s\in\mathbb{R}\). Then the Fourier multiplier operator \(e^{i\mu \left( D\right) }\) is bounded from \(W_{s}^{p,q}\left(\mathbb{R}^{n}\right) \) to \(W^{p,q}\left(\mathbb{R}^{n}\right) \) whenever \[ s>n\left( \alpha -2\right) \left| \frac{1}{p}-\frac{1}{2}\right| +n\left| \frac{1}{p}-\frac{1}{q}\right| . \] Theorem. Let \(\alpha \geq 2\) and \(\mu \) be a real-valued homogeneous function on \(\mathbb{R}^{n}\) of degree \(\alpha \) which belongs to \(C^{\infty }\left(\mathbb{R}^{n}\backslash \left\{ 0\right\} \right)\). Suppose there exists a point \( \xi _{0}\neq 0\) at which the Hessian determinant of \(\mu \) is not zero. Let \( \max \left( \frac{1}{q},\frac{1}{2}\right) \leq \frac{1}{p}\) or min\(\left( \frac{1}{q},\frac{1}{2}\right) \geq \frac{1}{p}\). Let \(s\in\mathbb{R}\) and suppose the Fourier multiplier operator \(e^{i\mu \left( D\right) }\) is bounded from \(W_{s}^{p,q}\left(\mathbb{R}^{n}\right) \) to \(W^{p,q}\left(\mathbb{R}^{n}\right)\). Then \[ s>n\left( \alpha -2\right) \left| \frac{1}{p}-\frac{1}{2}\right| +n\left| \frac{1}{p}-\frac{1}{q}\right| . \]