The Kakeya maximal operator on the variable Lebesgue spaces (Q481015): Difference between revisions

For every \(N\in\mathbb{N}\) denote by \(K_N\) the Kakeya maximal operator, i.e. the operator defined in the following way: \[ K_N(f)(x)=\sup_{R\in \mathcal{B}_N(x)} \frac{1}{|R|}\int_R |f|, \] where \(f\in L(\mathbb{R}^2), x\in\mathbb{R}^2\) and \(\mathcal{B}_N(x)\) denotes the set of all rectangles in the plane which contain \(x\) and have eccentricity \(N\). Let \(p:\mathbb{R}^2\rightarrow [1,\infty)\). a) It is said that the function \(p\) is locally log-Hölder continuous if there exists a positive constant \(c\) such that \[ |p(x)-p(y)|\log(1/|x-y|)\leq c, \;\;\;\;\;x,y\in\mathbb{R}^2,|x-y|<1; \] b) It is said that the function \(p\) is log-Hölder continuous at infinity if there exist positive constants \(c_\infty\) and \(p(\infty)\) such that \[ |p(x)-p(\infty)|\log(e+|x|)\leq c_\infty, \;\;\;\;\;x\in\mathbb{R}^2; \] c) Let \(p_{-}=\) ess inf \( p\) and \(p_{+}=\) ess sup \(p\); d) If \(1\leq p_{-}\leq p_{+}<\infty\) then for \(N\geq 1\) let \[ \beta_N=\sup\{|1/p(x)-1/p(y)|:x,y\in\mathbb{R}^2,|x-y|<\sqrt{N}\}. \] In works of A. Cordoba and J.-O. Stromberg the following estimate for norms of Kakeya maximal operators was proved: for any number \(p\in [2,\infty]\), \[ ||K_N||_{L^p(\mathbb{R}^2)\rightarrow L^p(\mathbb{R}^2)}\leq C (\log N)^{2/p}, \] where the constant \(C\) is independent of \(N\). In this article, the boundedness of Kakeya maximal operators on the variable Lebesgue spaces \(L^p(\mathbb{R}^2)\) is studied. In particular, the following result is established: let \(N\in \mathbb{N}\) and \(2\leq p_{-}\leq p_{+}<\infty\). Suppose that \(p\) is locally log-Hölder continuous and log-Hölder continuous at infinity. Then \[ ||K_N||_{L^p(\mathbb{R}^2)\rightarrow L^p(\mathbb{R}^2)}\leq C N^{p_{-}\beta_{N}}(\log N)^{2/p_{-}}, \] where the constant \(C\) is independent of \(N\). It is also established that if the function \(p\) is not constant, then the factor \(N^{p_{-}\beta_{N}}\) cannot be removed from the above estimate.