English

Identities and estimations involving the least common multiple of strong divisibility sequences

Number Theory 2020-04-13 v2

Abstract

In this paper, we first prove that for any strong divisibility sequences a=(an)n1\boldsymbol{a} = \left(a_n\right)_{n\geq 1}, we have the identity: lcm{(n0)a,(n1)a,,(nn)a}=lcm(a1,,an,an+1)an+1\mathrm{lcm} \left\lbrace \binom{n}{0}_{\bf{a}}, \binom{n}{1}_{\bf{a}},\dots, \binom{n}{n}_{\bf{a}} \right\rbrace = \frac{\mathrm{lcm} \left(a_1,\dots , a_n , a_{n+1}\right)}{a_{n+1}} (n1)\left(\forall n \geq 1\right), generalizing the identity of Farhi (obtained in 2009 for an=na_n=n). Then, we derive from this one some other interesting identities. Finally, we apply those identities to estimate the least common multiple of the consecutive terms of some Lucas sequences. Denoting by (Fn)n\left(F_n\right)_n the usual Fibonacci sequence, we prove for example that for all n1n \geq 1, we have Φn2494lcm(F1,,Fn)Φn23+4n3, \Phi^{\frac{n^2}{4}-\frac{9}{4}} \leq \mathrm{lcm}\left(F_1,\dots,F_n\right) \leq \Phi^{\frac{n^2}{3}+\frac{4n}{3}} , where Φ\Phi denotes the golden ratio.

Keywords

Cite

@article{arxiv.1907.06700,
  title  = {Identities and estimations involving the least common multiple of strong divisibility sequences},
  author = {Sid Ali Bousla and Bakir Farhi},
  journal= {arXiv preprint arXiv:1907.06700},
  year   = {2020}
}

Comments

11 pages