Abstract:We extend the Brauer-Siegel theorem to new families of number fields, both in the classical setting of asymptotically bad families and in the more general framework due to Tsfasman and Vlăduţ of asymptotically exact families. We introduce a notion of Galois complexity for extensions of number fields, and show that the generalized Brauer-Siegel theorem, as conjectured by Tsfasman and Vlăduţ, holds for families in which the complexity does not grow too fast. This allows to unify and extend all previously known results. The crucial step in our work is the proof of a new version -- stated in terms of our Galois complexity -- of a fundamental principle due to Stark descending exceptional zeroes of zeta functions down to quadratic number fields. Among the hitherto unknown cases we are able to treat are the families of number fields contained in the solvable Galois closure of a given number field.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is to extend the Brauer - Siegel theorem to new families of number fields, especially within the framework of asymptotically exact families proposed by Tsfasman and Vlăduţ. Specifically, the author introduces the concept of the Galois complexity of a number field extension and proves that if the Galois complexity of an asymptotically exact family does not grow too fast, then this family satisfies the generalized Brauer - Siegel theorem, as conjectured by Tsfasman and Vlăduţ. This enables the author to unify and extend all previously known results.
### Background and Motivation
The Brauer - Siegel theorem describes the asymptotic behavior of the product of the class number \(h(K)\) and the regulator of the unit group \(R(K)\) of a number field \(K\) with respect to the discriminant \(\Delta(K)\). This theorem was a central result in number theory in the 20th century and has a wide range of applications, for example, in sphere packing, class group index, geometric non - abelian theory, the André - Oort conjecture, and graph theory and public - key cryptosystems in computer science.
### Main Contributions
1. **Introduction of Galois Complexity**: For a number field extension \(L/K\), define its Galois complexity \(\xi_{L/K}\) as:
\[
\xi_{L/K}=\min\left\{\max_{1\leq i\leq s}[L_i : L_{i - 1}]_{\text{Gal}}, K = L_0\subset L_1\subset\cdots\subset L_s = L\right\}
\]
where \([L_i : L_{i - 1}]_{\text{Gal}}\) represents the degree of the Galois closure of \(L_i/L_{i - 1}\).
2. **Generalized Brauer - Siegel Theorem**: Proved that for any asymptotically exact family \((K_i)_{i\in\mathbb{N}}\), if it satisfies the condition:
\[
\lim_{i\rightarrow\infty}\frac{\log|\Delta(K_i)|}{\log\xi_{K_i/\mathbb{Q}}}=+\infty
\]
then this family satisfies the generalized Brauer - Siegel theorem, that is:
\[
\lim_{i\rightarrow\infty}\frac{\log(h(K_i)R(K_i))}{\log\sqrt{|\Delta(K_i)|}}=\beta(K)
\]
where \(\beta(K)\) is a real number defined by Tsfasman and Vlăduţ, ranging from \([2/5, 6/5]\).
3. **New Zero - Descent Principle**: Proved a new zero - descent principle, that is, if the zeta function of a number field extension \(L/K\) has a zero at some real number \(\rho\) and satisfies:
\[
1-(32g(L)\xi''_{L/K})^{-1}\leq\rho < 1
\]
then there exists a sub - extension \(F/K\) such that \([F:K]\leq 2\) and \(\zeta_F(\rho)=0\).
### Applications and Significance
- **Unify and Extend Existing Results**: The results of this paper cover the results previously obtained by Zykin, Dixit, Wong, etc. under specific conditions.
- **Handle a Wider Family of Number Fields**: In particular, it can handle the family of number fields contained in the solvable Galois closure of a given number field.
- **Improve Other Applications**: The new zero - descent principle is not only applicable to the Brauer - Siegel theorem, but can also improve other applications of Stark's results, such as the conclusions in Stark's Theorem 1 and 2.