Central oscillating multipliers on compact Lie groups (Q627476)

Let \(G\) be a connected, simply connected, compact semisimple Lie group of dimension \(n\) and rank \(m\). Let \(g\) be the Lie algebra of \(G\), and let \(\Omega\) be a sufficiently smooth function, which is also symmetric under the dual Weyl group, on the unit sphere of \(g^*\). In this paper the authors study the oscillating multiplier operator \[ T_{\Omega, \alpha, \beta} (f)(x) = K_{\Omega, \alpha, \beta} \ast f(x), \quad\alpha, \beta \geq 0, \] where \(K_{\Omega, \alpha, \beta}\) is a central kernel, \(f\) is a smooth function on \(G\) and \(\ast\) denotes convolution. If \(\Omega \equiv 1\), then \(T_{\Omega, \alpha, \beta}\) will be denoted by \(T_{\alpha, \beta}\). Let \(H^p(G)\) be the Hardy space of \(G\). The aim of this paper is to study the \(L^p\) boundedness of the operator \(T_{\Omega, \alpha, \beta}\). In particular, the authors prove the following theorem: Theorem 1. Assume \(0 < \beta < 1\) and \(\Omega\) is sufficiently smooth. {\parindent=5mm \begin{itemize}\item[1)]\(T_{\Omega, \alpha, \beta}\) is bounded on \(L^p(G)\) if and only if \(| \frac{1}{2} - \frac{1}{p} | \leq \frac{\alpha}{n\beta}\). \item[2)]\(T_{\Omega, \alpha, \beta}\) is bounded on \(L^1(G)\) if and only if \(\alpha > \frac{n\beta}{2}\). \item[3)]For \(\max\{ \frac{n}{n+1}, \frac{n-1}{n-1+\beta}\} < p \leq 1, T_{\Omega, \alpha, \beta}\) is bounded on \(H^p(G)\) if \[ \frac{1}{p} - \frac{1}{2} \leq \frac{\alpha}{\beta n}. \] \end{itemize}} The authors also prove: Theorem 2. Let \(\beta > 1\) and \(\Omega \in L^{\infty}\). \(T_{\Omega, \alpha, \beta}\) is bounded on \(H^p(G)\) for all \(0 < p < \infty\), if \[ \left| \frac{1}{2} - \frac{1}{p} \right| \leq \frac{\alpha}{n}. \] The following theorem is also proven: Theorem 3. If \(\beta =1\), then \(T_{\alpha, \beta}\) is bounded on \(L^p\) for \(1 \leq p\), provided \[ \left| \frac{1}{2} - \frac{1}{p} \right| < \frac{\alpha}{n-1}. \] Many technical results are given that give estimates for \(K_{\Omega, \alpha, \beta}\). These estimates are then used to prove the above theorems.