Limiting weak-type behavior of the centered Hardy-Littlewood maximal function of general measures on the positive real line (Q6542566)

Given two Radon measures \(\mu\), \(\nu\) on \(X\) (mostly \(X=\mathbb R_+\) in this note) and the centered maximal operator \(M_\mu\nu(x)=\sup\nu(B(x,r))/\mu(B(x,r))\), the authors study the existence and the value of the limit \(\lim_{\lambda\to 0+}\lambda\mu\) \((\{M_\mu\nu>\lambda\})\). When \(\mu=m_n\) is the Lebesgue measure on \(X=\mathbb R^n\), \textit{P. Janakiraman} [Trans. Am. Math. Soc. 358, No. 5, 1937--1952 (2006; Zbl 1096.42005)] has shown that this limit is \(\nu(\mathbb R^n)\), while subsequent case studies by others have revealed other limits of the form \(c_\mu \nu(X)\), where \(c_\mu\) is a constant depending on the measure \(\mu\).\N\NThe main result of the paper under review states that if the said limit is the same number \(c_0\) for every Dirac measure \(\mu=\delta_x\) on \(X=\mathbb R_+\), then this limit is \(c_0\nu(\mathbb R_+)\) for every Radon measure \(\nu\) on the same space. Examples are given to show that \(c_0\) may be any value in the interval \([0,2]\). A counterexample is constructed to show that the existence of the limit for \(\delta_0\) alone is not enough for the conclusion; indeed, the corresponding limit for another \(\delta_2\) may fail to exist.