Bombieri–Vinogradov theorem

Bombieri–Vinogradov theorem
Name	Bombieri–Vinogradov theorem
Field	Number theory
Published	1965–1967
Author	Enrico Bombieri; A. I. Vinogradov

Bombieri–Vinogradov theorem is a major result in analytic number theory connecting the distribution of prime numbers in arithmetic progressions with averaged forms of the Generalized Riemann Hypothesis. It provides strong "average" error bounds for primes in progression moduli up to nearly the square root of a large parameter, combining techniques from harmonic analysis, sieve theory, and exponential sums. The theorem has deep links to work by Hadamard, de la Vallée Poussin, Linnik, and Selberg and is foundational for many results in multiplicative number theory and computational investigations.

Statement

The theorem asserts that for a large parameter X and any fixed A>0, the sum over moduli q ≤ Q of the maximal discrepancy between the prime counting function π(X; q, a) and the expected X/(φ(q) log X) is small when Q is about X^{1/2} times a logarithmic factor. The precise formulation involves the von Mangoldt function Λ(n), the Chebyshev functions ψ(X; q, a), and Euler's totient function φ(q), giving an estimate of the form ∑_{q ≤ Q} max_{(a,q)=1} |ψ(X; q, a) − X/φ(q)| ≪ X / (log X)^A for Q ≤ X^{1/2} (log X)^{-B}, where B depends on A. This statement refines uniformity in the distribution of primes beyond classical results of Dirichlet and Hadamard–de la Vallée Poussin, and complements conditional results implied by the Generalized Riemann Hypothesis (GRH).

Historical background and development

Origins of the theorem lie in early twentieth-century work on prime distribution by Jacques Hadamard, Charles-Jean de la Vallée Poussin, and later contributions by Ivan Vinogradov, Atle Selberg, and Yakov G. Sinai. The development of sieve methods by Brun, Venkov, and Rosser informed mid-century advances, while Enrico Bombieri and A. I. Vinogradov synthesized techniques in the 1960s to produce the averaged form now named after them. The result was influenced by Pólya and Turán’s perspectives on mean-value theorems and by work on exponential sums by I. M. Vinogradov and Hans Rademacher. Subsequent elaborations drew on innovations from Hugh L. Montgomery, Robert Vaughan, Atle Selberg, D. A. Goldston, and Enrico Bombieri himself, shaping a rich interplay among analytic tools originally explored in the contexts of the Riemann zeta function, the Dirichlet L-function, and the Large Sieve.

Sketch of proof and methods

Proof strategies blend multiple strands: analytic properties of Dirichlet L-function, the large sieve inequality developed by Yu. V. Linnik and refined by Montgomery, and sieve-theoretic ideas associated with Brun and Selberg. A key step is establishing mean-square estimates for error terms across characters and moduli using the large sieve, orthogonality of characters modulo q, and estimates for exponential sums similar to those studied by I. M. Vinogradov. The argument navigates zero-density estimates for zeros of Dirichlet L-function and uses bilinear forms exploited by Robert Vaughan and Atle Selberg to control remainder terms. Bombieri’s contribution applied a sophisticated combination of these tools to achieve the effective averaging up to the X^{1/2} barrier.

Applications and consequences

The theorem has numerous consequences: it yields many unconditional corollaries in additive and multiplicative problems studied by Goldston, Granville, and Heath-Brown, informs results on gaps between primes connected to work by Erdős and Rankin, and supports distributional claims used by Hooley and Friedlander in sieve problems. It underlies combinatorial applications linked to Green, Tao, and Ziegler by providing averaged equidistribution in arithmetic progressions needed in transference arguments. In computational number theory, it guides heuristics employed in algorithms influenced by Miller and Wiles and impacts explicit estimates for prime-counting functions relevant to implementations by researchers at institutions such as Institut des Hautes Études Scientifiques and Princeton University.

Generalizations and related results

Strengthenings and analogues include the Elliott–Halberstam conjecture posed by P. D. T. A. Elliott and Heini Halberstam, which predicts distribution up to Q ≈ X^{1−ε}, and results on Barban–Davenport–Halberstam theorems developed by Barban, Davenport, and Halberstam. The Bombieri–Friedlander–Iwaniec theorem by Roger Heath-Brown and D. R. Heath-Brown and the work of H. Iwaniec extend aspects to primes represented by polynomials and to settings involving automorphic L-functions studied by Langlands and Gelbart. Zero-density results by Huxley and Jutila refine the landscape, while the development of the large sieve and dispersion methods by Bombieri, Davenport, and Montgomery continue to yield new variants and applications across modern analytic number theory.

Examples and numerical evidence

Numerical investigations compare ψ(X; q, a) to X/φ(q) for specific moduli q and residues a using computations by teams at Stanford University, Universität Bonn, and MIT. For moderate X (10^6–10^12) and Q up to about X^{1/2} divided by polylogarithmic factors, observed averages match the theorem’s bounds, in line with data compiled by researchers inspired by Oliveira e Silva and Bennett. Computational verification often relies on explicit formulas for primes and zero computations for Dirichlet L-function akin to methods developed in large-scale verifications of zeros of the Riemann zeta function, and informs ongoing efforts to test stronger conjectures such as the Elliott–Halberstam conjecture investigated by contemporaries including Goldston, Pintz, and Yıldırım.

Category:Analytic number theory