On generalized Thabit numbers ( p + 1 ) ​ p 𝔞 − 1 (p+1)p^{\mathfrak{a}}-1 in the k k -Lucas sequence

Herbert Batte, Florian Luca, Pantelimon Stănică · 2026

Read On generalized Thabit numbers ( p + 1 ) ​ p 𝔞 − 1 (p+1)p^{\mathfrak{a}}-1 in the k k -Lucas sequence by Herbert Batte, Florian Luca, Pantelimon Stănică (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2603-22878.txt.

Introduction

Batte, Luca, and Stănică solve a number theory equation linking two named families. One side holds generalized Thabit numbers of a set form. The other side holds terms of the k-Lucas sequence. That sequence starts with set terms, then sums the prior k terms each step. The prime in the Thabit form must be a Mersenne or Fermat prime. The authors find all solutions across the allowed ranges. The ranges cover the indices and exponents in the equation.

Catalog record

CREATORHerbert Batte, Florian Luca, Pantelimon Stănică
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE49716
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2603.22878

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

On generalized Thabit numbers ( p + 1 ) ​ p 𝔞 − 1 (p+1)p^{\mathfrak{a}}-1 in the k k -Lucas sequence Herbert Batte, Florian Luca, Pantelimon Stănică Abstract Let k ≥ 2 k\geq 2 and { L n ( k ) } n ≥ 2 − k \{L_{n}^{(k)}\}_{n\geq 2-k} be the sequence of k k -Lucas numbers whose first k k terms are 0 , … , 0 , 2 , 1 0,\ldots,0,2,1 and each term afterwards is the sum of the preceding k k terms. In this paper, we solve the Diophantine equation L n ( k ) = ( p + 1 ) ​ p 𝔞 − 1 L_{n}^{(k)}=(p+1)p^{\mathfrak{a}}-1 , for a Mersenne or Fermat prime p = 2 ℓ ± 1 p=2^{\ell}\pm 1 , and positive integers n ≥ 2 n\geq 2 , k ≥ 2 k\geq 2 , 𝔞 ≥ 1 \mathfrak{a}\geq 1 and ℓ ≥ 1 \ell\geq 1 . Keywords and phrases : Generalized Thabit numbers, k -Lucas numbers; Mersenne primes; Fermat primes; 2 -adic valuation; Linear forms in logarithms; Dujella-Pethő reduction methods; Diophantine equations. 2020 Mathematics Subject Classification : 11B39, 11D61, 11D45, 11J86. ∗ Corresponding author 1 Introduction 1.1 Background Recurrence sequences occupy a central place in modern Number theory, with their properties and Diophantine aspects attracting sustained research interest. The Fibonacci sequence and its companion Lucas sequence are among the most celebrated examples, and both admit natural higher-order generalizations. For an integer k\geq 2 , the k -Fibonacci sequence \{F_{n}^{(k)}\}_{n\in\mathbb{Z}} is defined by the linear recurrence F_{n}^{(k)}=F_{n-1}^{(k)}+F_{n-2}^{(k)}+\cdots+F_{n-k}^{(k)},\quad\text{for all}\ n\geq 2, subject to the initial conditions F_{2-k}^{(k)}=\cdots=F_{-1}^{(k)}=F_{0}^{(k)}=0 and F_{1}^{(k)}=1 . Setting k=2 recovers the classical Fibonacci sequence, and the family \{F_{n}^{(k)}\} has attracted considerable attention from algebraic and combinatorial perspectives alike. The companion sequence of k -Lucas numbers \{L_{n}^{(k)}\}_{n\in\mathbb{Z}} satisfies the same recurrence, L_{n}^{(k)}=L_{n-1}^{(k)}+L_{n-2}^{(k)}+\cdots+L_{n-k}^{(k)},\quad\text{for all}\ n\geq 2, but is distinguished by the initial values L_{2-k}^{(k)}=\cdots=L_{-1}^{(k)}=0,~L_{0}^{(k)}=2,~L_{1}^{(k)}=1 . The case k=2 yields the classical Lucas sequence, which has found applications across Number theory, cryptography, and primality testing. Both k - families exhibit a rich arithmetic structure that continues to motivate new investigations. For a prime p and an integer \mathfrak{a}\geq 1 , the expression \displaystyle(p+1)p^{\mathfrak{a}}-1, is a natural prime-base analogue of the classical Thabit-type number 3\cdot 2^{\mathfrak{a}}-1 . Indeed, when p=2 , we recover the Thabit numbers 3\cdot 2^{\mathfrak{a}}-1 , for an integer \mathfrak{a}\geq 1 . The identity \displaystyle(p+1)p^{\mathfrak{a}}=\left[(p+1)p^{\mathfrak{a}}-1\right]+1, shows that one is searching for terms in a linear recurrence for which the successor has a rigid multiplicative structure “ a prime power times its successor ”. This arithmetic structure is the basic input throughout the paper. The arithmetic study of special Number-theoretic sequences inside linear sequences has a long history. For example, Altassan and Alan proved in [ 1 ] that the only Mersenne numbers 2^{\mathfrak{a}}-1 , that intersect with the sequence of k -Lucas numbers are 1=L_{1}^{(k)} , 3=L_{2}^{(k)} and 7=L_{4}^{(2)} . In this paper, we study the Diophantine equation \displaystyle L_{n}^{(k)}=(p+1)p^{\mathfrak{a}}-1, (1.1) where p=2^{\ell}\pm 1 is a Mersenne or Fermat prime, and n\geq 2 , k\geq 2 , \mathfrak{a}\geq 1 , \ell\geq 1 are positive integers. Note that for n=0,1 , we have L_{n}^{(k)}=2,1 and the Diophantine equation ( 1.1 ) has no positive solutions p and \mathfrak{a} . 1.2 Main Result We prove the following result. Theorem 1.1 . Let p=2^{\ell}\pm 1 be a Mersenne or Fermat prime and n\geq 2 , k\geq 2 , \mathfrak{a}\geq 1 , \ell\geq 1 be positive integers. Then, the only solutions to the Diophantine equation \displaystyle L_{n}^{(k)}=(p+1)p^{\mathfrak{a}}-1, are \displaystyle(n,k,p,\mathfrak{a},\ell)\in\{(5,2,3,1,2),(7,2,5,1,2),(6,3,3,2,2)\}. These solutions in Theorem 1.1 are summarized in Table 1.1 below. n k p \mathfrak{a} \ell Type L_{n}^{(k)} 5 2 3=2^{2}-1=2^{1}+1 1 2 Mersenne/Fermat 11 6 3 3=2^{2}-1=2^{1}+1 2 2 Mersenne/Fermat 35 7 2 5=2^{2}+1 1 2 Fermat 29 Table 1.1: All solutions (n,k,p,\mathfrak{a},\ell) to ( 1.1 ) with p=2^{\ell}\pm 1 a Mersenne or Fermat prime. 2 Some useful formulas and inequalities involving L_{n}^{(k)} In this section we collect some useful formulas and inequalities concerning k -generalized Lucas numbers. First, it is known that the formula \displaystyle L_{n}^{(k)}=3\cdot 2^{n-2}\quad{\text{\rm holds~for}}\quad 2\leq n\leq k. Furthermore, L_{k+1}^{(k)}=3\cdot 2^{k-1}-2. Using induction, one proves easily that the inequality L_{n}^{(k)}\leq 3\cdot 2^{n-2}-2\qquad{\text{\rm holds~for~all}}\qquad n\geq k+1. (2.1) The k -generalized Lucas numbers can be expressed in terms of the k -generalized Fibonacci numbers via the formula L_{n}^{(k)}=2F_{n+1}^{(k)}-F_{n}^{(k)}. (2.2) To prove the above formula, it is enough to first notice that it holds for n=-(k-2),\ldots,0,1,2 and then to invoke the fact that both sequences \{L_{n}^{(k)}\}_{n\in{\mathbb{Z}}} and \{2F_{n+1}^{(k)}-F_{n}^{(k)}\}_{n\in{\mathbb{Z}}} satisfy the same k th order linear recurrence to conclude that the above formula holds for all n\in{\mathbb{Z}} . Another useful formula is L_{n}^{(k)}=2L_{n-1}^{(k)}-L_{n-(k+1)}^{(k)} for all n\in{\mathbb{Z}} (to see why it holds, replace the left-hand side above by L_{n-1}^{(k)}+L_{n-2}^{(k)}+\cdots+L_{n-k}^{(k)} , cancel a term L_{n-1}^{(k)} on both sides of the resulting equation, and recognise that the resulting formula is just the formula representing L_{n-1}^{(k)} as the sum of the previous k terms of the sequence). Reducing the above formula modulo 2 we get L_{n}^{(k)}\equiv L_{n-(k+1)}^{(k)}\pmod{2}, showing that \{L_{n}^{(k)}\}_{n\in{\mathbb{Z}}} is periodic modulo 2 with period k+1 . This fact is so useful that we record it as a lemma. Lemma 2.1 . The sequence \{L_{n}^{(k)}\}_{n\in{\mathbb{Z}}} is periodic modulo 2 with period k+1 . Now let us recall some more formulas related to the sequence of k -generalized Lucas numbers. Its characteristic polynomial g_{k}(X)=X^{k}-X^{k-1}-\cdots-X-1, is irreducible in \mathbb{Q}[X] . This polynomial has exactly one real root greater than 1, denoted by \alpha:=\alpha(k) , and all its other roots lie inside the unit circle in the complex plane. Further, this real root lies within the range \displaystyle 2\left(1-2^{-k}\right)<\alpha<2,\qquad{\text{\rm for~all}}\qquad k\geq 2. (2.3) If we need to refer to all the roots of g_{k}(X) , then we label them as \alpha_{1},\ldots,\alpha_{k} with the convention that \alpha:=\alpha_{1} . For all n\geq 0 and k\geq 2 , it was proved in [ 4 ] that the k -generalized Lucas numbers are bounded by powers of \alpha as \displaystyle\alpha^{n-1}\leq L_{n}^{(k)}\leq 2\alpha^{n}. (2.4) Moreover, if we compare ( 1.1 ) and ( 2.4 ), we deduce that \displaystyle p^{2}\leq p^{\mathfrak{a}+1}=p\cdot p^{\mathfrak{a}}<(p+1)p^{\mathfrak{a}}=1+L_{n}^{(k)}\leq 3\alpha^{n}, which implies that \displaystyle p<\sqrt{3}\alpha^{n/2}\qquad\text{and}\qquad\mathfrak{a}\leq n. (2.5) Next, we define the function f_{k}(x):=\frac{x-1}{2+(k+1)(x-2)},\qquad x\in{\mathbb{C}}\backslash\left\{2-{2}/{(k+1)}\right\}. We have \frac{df_{k}(x)}{dx}=-\frac{(k-1)}{(2+(k+1)(x-2))^{2}}<0,\qquad{\text{\rm for all}}\qquad x>0. In particular, inequality ( 2.3 ) implies that \displaystyle\dfrac{1}{2}=f_{k}(2)<f_{k}(\alpha)<f_{k}(2(1-2^{-k}))\leq\dfrac{3}{4}, (2.6) for all k\geq 3 . It is easy to check that the above inequality holds for k=2 as well. Further, it is easy to verify that |f_{k}(\alpha_{i})|<1 , for all 2\leq i\leq k , where \alpha_{i} are the remaining roots of \Psi_{k}(x) for i=2,\ldots,k . The Binet formula for the general term of the k -Lucas numbers is given by L_{n}^{(k)}=\sum_{i=1}^{k}(2\alpha_{i}-1)f_{k}(\alpha_{i})\alpha_{i}^{n-1},\qquad{\text{\rm for all}}\quad n\in{\mathbb{Z}}. Since \alpha_{2},\ldots,\alpha_{k} are inside the unit circle, one would say that only the first term (2\alpha_{1}-1)f_{k}(\alpha_{1})\alpha_{1}^{n-1} already approximates L_{n}^{(k)} and indeed the approximation \displaystyle\left|L_{n}^{(k)}-f_{k}(\alpha)(2\alpha-1)\alpha^{n-1}\right|<\frac{3}{2},\qquad{\text{\rm for~all}}\quad n\geq 2-k (2.7) appears, for example, in [ 4 ] . This tells us that most of the size of L_{n}^{(k)} comes from the dominant term involving the real root \alpha , while the other terms contribute very little. A better estimate than ( 2.7 ) appears in Section 3.3 page 14 of [ 2 ] , but with a more restricted range of n in terms of k . It states that \displaystyle\left|f_{k}(\alpha)(2\alpha-1)\alpha^{n-1}-3\cdot 2^{n-2}\right|<3\cdot

Read the complete text