On the number of unitary divisors of integers in arithmetic progressions (Q1775065)

Let \(1\leq l\leq k\) and \((k,l)= 1\), and define \[ d_{k,l}(n)= \prod_{p^\alpha\| n,p\equiv l\pmod k}(\alpha+ 1), \;d^*_{k,l}(n)= \prod_{p^\alpha\| n,p\equiv l\pmod k}2, \] so \(d^*_{1,1}(n)\) is the number of unitary divisors of \(n\). Let \[ F(n; k,l)= {\ln(d_{k,l}(n))\ln(\varphi(k)\ln n)\over \ln n}, \] \(F^*(n; k,l)\) be the corresponding function with \(d^*_{k,l}(n)\) instead of \(d_{k,l}(n)\), and \(D^*(n;k,l)= F(n; k,l)- F^*(n; k,l)\). In [C. R., Math., Acad. Sci. Paris 339, No. 2, 87--90 (2004; Zbl 1119.11057)], \textit{A. Derbal} and \textit{A. Smati} showed that \(\limsup F(n;k,l)= \ln 2\) and that \(F(n;k,l)\) reaches its maximum among the \(d_{k,l}(n)\)-superior highly composite numbers. In this paper, the author investigates the analogous problems for \(F^*(n;k,l)\) and \(D^*(n;k,l)\). Reviewer's remark: The reviewer noted that the definitions of \(d_{k,l}(n)\) and \(d_{k,l}(n)\) in the abstract of this paper differ from those given above and in the text, and that a similar remark applies to the definition of \(d_{k,l}(n)\) in the paper by A. Derbal and A. Smati quoted above.