A refinement of the Burgess bound for character sums (Q2195006): Difference between revisions

For a prime number \(q\) and a multiplicative character \(\chi\) modulo \(q\), the authors considered giving a refinement for bound of the sums \(\sum_{M<n\leq M+N}\chi(n)\). The first result in this direction, initiated by \textit{G. Pólya} [Nachr. Ges. Wiss. Göttingen, Math.-Phys. Kl. 1918, 21--29 (1918; JFM 46.0265.02)] and \textit{I. M. Vinogradov} [J. Soc. Phys. Math. Univ. Perm 1, 18--28 (1918; JFM 48.1352.04)] is of the form \[ \sum_{M<n\leq M+N}\chi(n)=O(q^{1/2}\log q). \] Several logarithmic improvements have been developed thereafter, the sharpest bound for which has been due to \textit{D. A. Burgess}, [Proc. Lond. Math. Soc. 12, 193--206 (1962; Zbl 0106.04004); Proc. Lond. Math. Soc. 13, 524--536 (1963; Zbl 0123.04404)], stated as \[ \sum_{M<n\leq M+N}\chi(n)=O(N^{1-1/r}q^{(r+1)/4r^2}\log q). \] Slight refinements of the Burgess bound mentioned above have been made by a factor of \((\log q)^{1/r}\) and \((\log q)^{1/2r}\). In this paper under review, the authors give a refinement of the Burgess bound for any integer \(r\geq 2\), by a factor of \((\log q)^{1/4r}\), as follows \[ \sum_{M<n\leq M+N}\chi(n)=O(N^{1-1/r}q^{(r+1)/4r^2}(\log q)^{1/{4r}}). \] The improvement above is due to averaging over numbers with no small prime factors rather than over an entire interval, an idea elaborated from previous approaches.