On an additive prime divisor function of Alladi and Erdos

Dorian Goldfeld
DOI: https://doi.org/10.48550/arXiv.1605.09044
2016-06-18
Abstract:This paper discusses the additive prime divisor function $A(n) := \sum\limits_{p^\alpha || n} \alpha \, p$ which was introduced by Alladi and Erd\H os in 1977. It is shown that $A(n)$ is uniformly distributed (mod $q$) for any fixed integer $q > 1.$
Number Theory
What problem does this paper attempt to address?
The problem that this paper attempts to solve is about the uniform distribution property of the additive prime divisor function \(A(n)\). Specifically, the paper focuses on the additive function \(A(n)\) introduced by Alladi and Erdős in 1977, which is defined as: \[ A(n) := \sum_{i = 1}^{r} a_i \cdot p_i \] where \(n = p_1^{a_1}p_2^{a_2}\cdots p_r^{a_r}\) is the unique prime factorization form of the positive integer \(n\). Alladi and Erdős have already proven that \(A(n)\) is uniformly distributed modulo 2. This paper further explores the uniform distribution property of \(A(n)\) modulo any fixed integer \(q>1\) and gives an explicit bound for the error term. The specific goal is to prove that for any fixed relatively prime integers \(h\) and \(q\) (where \(q > 2\)), as \(x\rightarrow\infty\), we have: \[ \sum_{n\leq x}e^{2\pi i hA(n)/q}=\begin{cases} C_{h,q}\cdot\frac{x}{(\log x)^{1+\mu(q)/\phi(q)}}\left(1 + O\left(\frac{1}{\log x}\right)\right)&\text{if }\mu(q)\neq0,\\ O\left(x e^{-c_0\sqrt{\log x}}\right)&\text{if }\mu(q) = 0, \end{cases} \] where \(C_{h,q}\) is a constant depending on \(h\) and \(q\), and \(\mu(q)\) and \(\phi(q)\) are the Möbius function and the Euler function respectively. Furthermore, the paper also proves the uniform distribution theorem of \(A(n)\) modulo \(q\), that is: \[ \sum_{n\leq x, A(n)\equiv h\pmod{q}}1=\frac{x}{q}+O\left(x\sqrt{\log x}\right) \] This result shows that as \(x\) tends to infinity, the number of positive integers \(n\) satisfying \(A(n)\equiv h\pmod{q}\) is proportional to \(x\), and the error term is \(O\left(x\sqrt{\log x}\right)\).