scientific article; zbMATH DE number 19859

From MaRDI portal
Publication:3979369





[For the entire collection see Zbl 0727.00003.] Let \(\sigma_ \alpha=\sum_{d\mid n}d^ \alpha\). The authors prove for \(0<\alpha<3/2\) the following asymptotic formula \[ \begin{aligned} \sum_{n\leq T}\sigma_ 0(n)\sigma_ \alpha(n+N)& = C_ 1(N,\alpha)T^{1+\alpha}\log T+C_ 2(N,\alpha)T^{1+\alpha}+C_ 3(N,\alpha)T\log T+C_ 4(N,\alpha)T \\ & +O(T^{{2(1+\alpha)^ 2 \over 3+2\alpha}}\log^ 3(T)N^{{1\over 6+4\alpha}}\sigma_ 0(N)\log(N+1)) \end{aligned} \] with some explicitly given constants \(C_ i(N,\alpha)\), which are too involved to be given here. The proof depends on the analytic continuation of \textit{N. V. Kuznetsov's} convolution formula in [J. Sov. Math. 29, 1131-1159 (1985), translation from Zap. Nauchn. Semin. Leningr. Otd. Mat. Inst. Steklova 129, 43-84 (1983; Zbl 0496.00013)]. They claim that Kuznetsov's asymptotic formula in loc. cit. for \(\sum_{n\leq T}\sigma_ 0(n)\sigma_ 0(n+N)\) can be recovered from their's by letting \(\alpha\) tend to zero.











This page was built for publication:

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3979369)