Zero-Free Regions for Dirichlet L-Functions, and the Least Prime in an Arithmetic Progression (Q4711129)
From MaRDI portal
!
This is the item page for this Wikibase entity, intended for internal use and editing purposes. Please use the normal view instead:
scientific article; zbMATH DE number 1264
| Language | Label | Description | Also known as |
|---|---|---|---|
| default for all languages | No label defined |
||
| English | Zero-Free Regions for Dirichlet L-Functions, and the Least Prime in an Arithmetic Progression |
scientific article; zbMATH DE number 1264 |
Statements
Zero-Free Regions for Dirichlet L-Functions, and the Least Prime in an Arithmetic Progression (English)
0 references
25 June 1992
0 references
zero-free regions for Dirichlet \(L\)-functions
0 references
Linnik's constant
0 references
1.0000002
0 references
0.8745507
0 references
0.83485186
0 references
0.79007816
0 references
0.7749301
0 references
0.7748362
0 references
0.7687392
0 references
0.75097406
0 references
The two main purposes of this paper are to give new zero-free regions for Dirichlet \(L\)-functions, and to give a new estimate for Linnik's constant. If \(q\) is sufficiently large it is shown that the \(L\)-functions to modulus \(q\) have at most a Siegel zero in the region \(\sigma \geq 1-0.348/\log q\), \(| t| \leq 1\). The previous best result [see \textit{J.-R. Chen}, Sci. Sin., Ser. A 26, 714--731 (1983; Zbl 0513.10045)] had a constant \(0.103\). Moreover it is shown that there is at most one character \(\chi_ 1\), together with its conjugate, for which the \(L\)-function \(L(s,\chi\)) can have a zero in the region \(\sigma \geq 1-0.702/\log q\), \(| t| \leq 1\); and there are at most two characters \(\chi_ 1,\chi_ 2\), together with their conjugates, which can have a zero in the region \(\sigma \geq 1-0.857/\log q\), \(| t| \leq 1.\) Using these results it is shown that if \((a,q)=1\) then there is a prime \(p\equiv a\pmod q\) satisfying \(p\ll q^{11/2}\). The previous best bound for Linnik's constant [\textit{J.-R. Chen} and \textit{J.-M. Liu}, Sci. China, Ser. A 32, 654--673, 792--807 (1989; Zbl 0684.10040)] had \(27/2\) in place of \(11/2\). The proofs of these results depend on a number of new ideas, of which we mention three. Firstly it is shown how estimates for zero-free regions can be made to depend on Burgess's bounds for \(L\)-functions. In particular, his recent work [\textit{D. A. Burgess}, J. Lond. Math. Soc., II. Ser. 33, 219--226 (1986; Zbl 0593.10033)] on the ``\(r=3\)'' case is used. Secondly, the partial fraction decomposition for \(\frac{L'}{L}(s,\chi)\) which has previously been employed in obtaining zero-free regions, is replaced by an ``explicit formula'' for a sum \(\sum \Lambda (n)\chi (n)n^{-s}w(n)\) with a general non-negative weight \(w(n)\). Thirdly, the zero-density estimate used by other authors is augmented by a new kind of density estimate, the origins of which are related to the apparatus for proving zero-free regions. It should also be mentioned that the paper requires a considerable amount of numerical work -- much more than in previous works on the subject. The paper concludes by describing a number of ways in which small improvements to the results may be obtained.
0 references