Distribution of sums of square roots modulo $1$

Siddharth Iyer

2024-04-01

Abstract:We improve upon a result of Steinerberger (2024) by demonstrating that for any fixed $k \in \mathbb{N}$ and sufficiently large $n$, there exist integers $1 \leq a_1, \dots, a_k \leq n$ satisfying: \begin{align*} 0 < \left\| \sum_{j=1}^{k} \sqrt{a_j} \right\| = O(n^{-k/2}). \end{align*} The exponent $k/2$ improves upon the previous exponent of $c k^{1/3}$ of Steinerberger (2024), where $c>0$ is an absolute constant. We also show that for $\alpha \in \mathbb{R}$, there exist integers $1 \leq b_1, \dots, b_k \leq n$ such that: \begin{align*} \left\| \sum_{j=1}^k \sqrt{b_j} - \alpha \right\| = O(n^{-\gamma_k}), \end{align*} where $\gamma_k \geq \frac{k-1}{4}$ and $\gamma_k = k/2$ when $k=2^m - 1$, $m=1,2,\dots$. Importantly, our approach avoids the use of exponential sums.

Number Theory

What problem does this paper attempt to address?

The problem that this paper attempts to solve is related to square roots and the distribution under modulo 1. Specifically, the author aims to improve the results of Steinerberger (2024) and shows that for any fixed \(k\in\mathbb{N}\) and sufficiently large \(n\), there exist integers \(1\leq a_1,\ldots,a_k\leq n\) such that:

\[0 < \left\|\sum_{j = 1}^k\sqrt{a_j}\right\|=O(n^{-k/2})\]

Here, \(\|x\|\) represents the distance of \(x\) from the nearest integer, that is, \(\|x\|=\min_{m\in\mathbb{Z}}|x - m|\).

Furthermore, the author also proves that for any real number \(\alpha\), there exist integers \(1\leq b_1,\ldots,b_k\leq n\) such that:

\[\left\|\sum_{j = 1}^k\frac{a}{b_j}-\alpha\right\|=O(n^{-\gamma_k})\]

where \(\gamma_k\geq\frac{k - 1}{4}\), and when \(k = 2m-1\), \(\gamma_k=\frac{k}{2}\).

### Main contributions

1. **Improved exponent**: The author improves Steinerberger (2024)'s exponent \(ck^{1/3}\) to \(k/2\), providing a stronger result.
2. **Avoiding exponential sums**: Different from Steinerberger's method, the author's method does not rely on exponential sums but is achieved through other technical means.
3. **Wider applicability**: It not only considers the distribution of square root sums but also studies the distribution of weighted sums and gives more precise bounds.

### Conclusion

This paper provides a stronger and more explicit result by conducting an in - depth analysis of the distribution of square root sums under modulo 1, while avoiding the complex exponential sum techniques in traditional methods. This provides new ideas and tools for further research in related fields.