English

Nontrivial upper bounds for the least common multiple of an arithmetic progression

Number Theory 2020-04-17 v1

Abstract

In this paper, we establish some nontrivial and effective upper bounds for the least common multiple of consecutive terms of a finite arithmetic progression. Precisely, we prove that for any two coprime positive integers aa and bb, with b2b\geq 2, we have lcm(a,a+b,,a+nb)(c1blogb)n+ab    (nb+1),\mathrm{lcm}\left(a,a+b,\dots,a+nb\right) \leq \left(c_1\cdot b\log b\right)^{n+\left\lfloor \frac{a}{b}\right\rfloor}~~~~(\forall n\geq b+1), where c1=41.30142c_1=41.30142. If in addition bb is a prime number and a<ba<b, then we prove that for any nb+1n\geq b+1, we have lcm(a,a+b,,a+nb)(c2bbb1)n\mathrm{lcm}\left(a,a+b,\dots,a+nb\right) \leq \left(c_2\cdot b^{\frac{b}{b-1}}\right)^n, where c2=12.30641c_2=12.30641. Finally, we apply those inequalities to estimate the arithmetic function MM defined by M(n):=1φ(n)1nn=11M(n):=\frac{1}{\varphi(n)}\sum_{\substack{1\leq\ell\leq n \\ \ell \wedge n=1}}\frac{1}{\ell} (n1\forall n \geq 1), as well as some values of the generalized Chebyshev function θ(x;k,)\theta(x;k,\ell).

Keywords

Cite

@article{arxiv.2004.07335,
  title  = {Nontrivial upper bounds for the least common multiple of an arithmetic progression},
  author = {Sid Ali Bousla},
  journal= {arXiv preprint arXiv:2004.07335},
  year   = {2020}
}

Comments

8 pages