Weighted norm inequalities for maximal convolution-type operators (Q1974431)

The purpose here is the study of convergence questions for more general convolution type operators as \[ (T_\delta f)(x)= \int_{\mathbb{R} ^n} \phi_\delta(x,t)f(t) d\nu(t) \] where \(d\nu\) is a measure on \(\mathbb{R} ^n\), \(\phi_\delta\) is a nonnegative measurable function defined on \(\mathbb{R} ^n\times\mathbb{R} ^n\) and \(\delta\) belongs to an arbitrary collection \(\Gamma\) of measurable functions. Conditions on the measures \(d\mu\) and \(d\nu\) are found by the author in order that the sequence \(((T_{\delta_{ix}}f)(x))_{i\geq 1}\) converges, with respect to the measure \(d\mu\), pointwise (see Theorems 4.2 and 4.3) and in norm (see Theorems 5.2 and 5.3). This is done whenever \((\delta_{ix})_{i\geq 1}\) is a suitable subsequence of \(\Gamma\). The proof of the pointwise convergence is obtained by means of a weak-type norm inequality for an operator related to a maximal function previously introduced by \textit{R. L. Bagby} [Rocky Mt. J. Math. 11, 243-260 (1981; Zbl 0458.42012)]. And the norm convergence makes use of a strong-type inequality for a suitable operator which is introduced and studied by the author in section 2.