Abstract:We derive and implement a new way to find lower bounds on the smallest limiting trace-to-degree ratio of totally positive algebraic integers and improve the previously best known bound to 1.80203. Our method adds new constraints to Smyth's linear programming method to decrease the number of variables required in the new problem of interest. This allows for faster convergence recovering Schur's bound in the simplest case and Siegel's bound in the second simplest case of our new family of bounds. We also prove the existence of a unique optimal solution to our newly phrased problem and express the optimal solution in terms of polynomials. Lastly, we solve this new problem numerically with a gradient descent algorithm to attain the new bound 1.80203.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is to improve the lower - bound estimation of the trace - to - degree ratio of totally positive algebraic integers. Specifically, the authors improved Smyth's linear programming method by introducing new constraint conditions, thereby reducing the number of variables required in the problem, achieving a faster convergence rate, and finally obtaining a new lower bound of 1.80203.
### Background and Problem Description
For an algebraic integer \(\alpha\), its trace, norm, and discriminant are defined as follows respectively:
\[
\text{tr}(\alpha) := \frac{1}{n} \sum_{i = 1}^n \alpha_i,
\]
\[
\text{Nr}(\alpha)=\prod_{i = 1}^n \alpha_i,
\]
\[
\Delta(\alpha)=\prod_{i < j}(\alpha_i-\alpha_j)^2\in\mathbb{Z}^+.
\]
If all Galois conjugates \(\alpha_i>0\), then \(\alpha\) is said to be totally positive. Denote \(A\) as the set of all totally positive algebraic integers. The Schur - Siegel - Smyth trace problem aims to calculate:
\[
\lambda_{\text{SSS}}:=\liminf_{\alpha\in A}\text{tr}(\alpha).
\]
### Historical Progress
- Schur (1918) proved that \(\lambda_{\text{SSS}}\geq\sqrt{e}\).
- Siegel (1945) improved this result and proved that:
\[
\lambda_{\text{SSS}}\geq e(1 + \nu^{-1})^{-\nu}\approx1.7336105,
\]
where \(\nu\) is the positive root of the transcendental equation
\[
(1 + \nu)\log(1 + \nu^{-1})+\log\nu/(1 + \nu)=1.
\]
### New Method and Contribution
The authors improved Smyth's linear programming method by introducing new constraint conditions. The specific steps include:
1. **Introducing New Constraints**: Add new constraint conditions on the basis of Smyth's method to reduce the number of variables required.
2. **Optimization Problem Reconstruction**: Reconstruct the problem into a new optimization problem and prove that there is a unique optimal solution.
3. **Numerical Solution**: Use the gradient descent algorithm to numerically solve the new problem and obtain a new lower bound of 1.80203.
### Main Results
Through these improvements, the authors not only increased the estimated value of the lower bound but also showed how to achieve better results by introducing fewer polynomials. For example, Wang, Wu, and Wu used 130 polynomials to reach a lower bound of 1.7931, while this paper only used 4 polynomials to reach a lower bound of 1.80203.
### Summary
The main contribution of this paper lies in significantly improving the lower - bound estimation of the trace - to - degree ratio of totally positive algebraic integers by introducing new mathematical tools and constraint conditions, providing new ideas and methods for further research in related fields.