On an asymmetric divisor problem with congruence conditions (Q1121318): Difference between revisions

The authors obtain \(O\)- and \(\Omega\)-results for the error terms in asymptotic formulas involving certain divisor functions. The functions in question are \(d_b(\ell,k;n)\) (the number of \(m\in\mathbb N\) such that \(m^b\mid n\) and \(m\equiv \ell \pmod k\) for \(b\geq 2\) a fixed integer), \(d_{a,b}(\ell_1,k_1;\ell_2,k_2;n)\) (the number of \((v,w)\in\mathbb N^2\) with \(v^aw^b=n\), \(v\equiv \ell_1\pmod{k_1}\), \(w\equiv \ell_2\pmod{k_2}\), \(\ell_ i\leq k_i\), \(a\neq b\) fixed natural numbers) and \(d_{a,b}(\ell,k;n)\) (the number of \((v,w)\in\mathbb N^2\) such that \(v^aw^b=n\), \(w\equiv \ell \pmod k\), \(\ell \leq k\), \(a\neq b\) fixed natural numbers). The proofs of the (uniform) \(O\)-estimates, which are precisely stated in the text, depend on explicit formulas for the error terms involving the function \(\Psi(t)=t-[t]\) and the use of (one-dimensional) exponent pairs. In particular, the exponent pair \((9/56+\varepsilon, 37/56+\varepsilon)\), obtained recently by \textit{M. N. Huxley} and \textit{N. Watt} [Proc. Lond. Math. Soc., III. Ser. 57, No. 1, 1--24 (1988; Zbl 0644.10027)], is extensively used.