An Unexpected Connection Between the Discrete Zeta Function and the Erdős–Straus Conjecture Under Mballa’s Conjecture

Philemon Urbain Mballa · 2025

Read An Unexpected Connection Between the Discrete Zeta Function and the Erdős–Straus Conjecture Under Mballa’s Conjecture by Philemon Urbain Mballa (2025) 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-2505-08716.txt.

Introduction

Philemon Urbain Mballa links a classic unsolved problem about unit fractions to the behavior of the zeta function. The Erdős–Straus conjecture claims every fraction 4/n splits into three unit fractions; it has resisted proof for decades. His 2025 paper works under a personal conjecture, stated in an earlier article, that parametrizes solutions of that splitting problem. From those solutions he builds a series whose terms are triples of unit fractions. Summed, the series equals a simple transform of the discrete zeta function.

The bridge runs both ways in spirit: facts about zeta values meet facts about Egyptian fractions. Because the key step rests on his own unproven conjecture, the connection is conditional. The author presents it openly as such, and offers the bridge as a route others may test.

Catalog record

CREATORPhilemon Urbain Mballa
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE16 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2505-08716

Browse the full library · About the Tower of Babel

Full text

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

An Unexpected Connection Between the Discrete Zeta Function and the Erdős–Straus Conjecture Under Mballa’s Conjecture Philemon Urbain Mballa Abstract In this article, we establish an additive decomposition of the discrete zeta function (for s ∈ ℕ ∗ , s > 1 s\in\mathbb{N}^{*},\ s>1 ), more precisely of the function 4 ​ ( ζ ​ ( s ) − 1 ) 4(\zeta(s)-1) , as a series whose general term is of the form 1 x n ​ ( s ) + 1 y n ​ ( s ) + 1 z n ​ ( s ) \frac{1}{x_{n}(s)}+\frac{1}{y_{n}(s)}+\frac{1}{z_{n}(s)} , where x n ​ ( s ) , y n ​ ( s ) , z n ​ ( s ) x_{n}(s),y_{n}(s),z_{n}(s) are solutions of the Erdős–Straus conjecture under a personal conjecture (which I will refer to here as Mballa’s Conjecture) that I formulated by parametrization in the article: http://arxiv.org/abs/2502.20935 . This connection thus builds a bridge between analysis and Egyptian fractions in general and the Erdős conjecture in particular. 1 Introduction The Erdős–Straus conjecture posits that for all integers n\geq 2 , the rational number \frac{4}{n} can be written as the sum of three unit fractions: \frac{4}{n}=\frac{1}{x}+\frac{1}{y}+\frac{1}{z},\quad x,y,z\in\mathbb{N}^{*}. This equation is equivalent to \frac{4x-n}{nx}=\frac{y+z}{yz}. In order for the left-hand side to belong to \mathbb{N}^{*} , it is necessary that x\geq\left\lfloor\frac{n}{4}\right\rfloor+1, where \left\lfloor x\right\rfloor denotes the integer part of x , and this notation will be used throughout the rest of the article. In this work, we propose a parameterization of the solutions, which allows us to define a discrete zeta-type function whose terms are derived directly from the decomposition \frac{4}{n^{s}}=\frac{1}{x_{n}(s)}+\frac{1}{y_{n}(s)}+\frac{1}{z_{n}(s)} under Mballa’s conjecture. This leads to a functional identity that coincides with 4(\zeta(s)-1) , and opens a new analytic approach to the problem. 2 Mballa’s Conjecture Mballa’s Conjecture posits that For every n\in\mathbb{N} with n\geq 2 , there exist (x,t)\in(\mathbb{N}^{*})^{2} such that t^{2}(4x-n)^{2}-2nxt=q^{2},\quad\text{with }q\in\mathbb{N}. This conjecture has been numerically verified in a previous article up to a reasonable bound on n (see http://arxiv.org/abs/2502.20935 for more details). In the appendix of this paper, we provide a code that will allow the reader, if desired, to test Mballa’s Conjecture numerically. In the paper http://arxiv.org/abs/2502.20935 , we established that for all n\in\mathbb{N} , with n\geq 2 , if there exists x\in\mathbb{N}^{*} such that x\geq\left\lfloor\frac{n}{4}\right\rfloor+1 , and t\in\mathbb{N}^{*} , satisfying t^{2}(4x-n)^{2}-2nxt=q^{2},\quad\text{with }q\in\mathbb{N}, then we can construct y=t(4x-n)-q,\quad z=t(4x-n)+q\in\mathbb{N}^{*} such that \frac{4}{n}=\frac{1}{x}+\frac{1}{y}+\frac{1}{z}. It follows from the above that Mballa’s Conjecture implies the Erdős–Straus Conjecture. 3 Construction of the Functional Identity For each natural number n\geq 2 , let s\in\mathbb{N} be fixed with s>1 , so that n^{s}\in\mathbb{N} . Under Mballa’s Conjecture, there exist (x_{n}(s),y_{n}(s),z_{n}(s))\in(\mathbb{N}^{*})^{3} such that \frac{4}{n^{s}}=\frac{1}{x_{n}(s)}+\frac{1}{y_{n}(s)}+\frac{1}{z_{n}(s)}, with x_{n}(s)>\left\lfloor\frac{n^{s}}{4}\right\rfloor+1,\quad y_{n}(s)=t(4x_{n}(s)-n^{s})-q_{n}(s),\quad z_{n}(s)=t(4x_{n}(s)-n^{s})+q_{n}(s), where q_{n}(s)\in\mathbb{N} satisfies q_{n}(s)^{2}=t^{2}(4x_{n}(s)-n^{s})^{2}-2tx_{n}(s)n^{s}. Let us consider the following series, which we denote here by \zeta_{M}(s) : \zeta_{M}(s):=\sum_{n=2}^{\infty}\left(\frac{1}{x_{n}(s)}+\frac{1}{y_{n}(s)}+\frac{1}{z_{n}(s)}\right), where x_{n}(s),y_{n}(s),z_{n}(s)\in\mathbb{N}^{*} are constructed from the decomposition of \frac{4}{n^{s}} under the conjecture described above. Starting from the equality above and summing over n , we obtain: \sum_{n=2}^{\infty}\frac{4}{n^{s}}=4(\zeta(s)-1)=\zeta_{M}(s):=\sum_{n=2}^{\infty}\left(\frac{1}{x_{n}(s)}+\frac{1}{y_{n}(s)}+\frac{1}{z_{n}(s)}\right), which defines a functional identity between the discrete series \zeta_{M}(s) and the classical discrete Riemann zeta function. 4 Convergence of \zeta_{M}(s) 4.1 Method 1: Direct Equivalence We can clearly see that the general term \frac{1}{x_{n}(s)}+\frac{1}{t(4x_{n}(s)-n^{s})-q_{n}(s)}+\frac{1}{t(4x_{n}(s)-n^{s})+q_{n}(s)}=\frac{4}{n^{s}}, with q_{n}(s) satisfying t^{2}(4x_{n}(s)-n^{s})^{2}-2tn^{s}x_{n}(s)=q_{n}(s)^{2}. Now, the term \frac{4}{n^{s}} belongs to a multiplicatively adjusted Riemann-type series, which converges for s>1 . Therefore, the series \zeta_{M}(s) converges. 4.2 Method 2: Comparison Theorem Proposition (Linearity of Series). Let (u_{n}) and (v_{n}) be two sequences with values in a field \mathbb{K} , and let \lambda\in\mathbb{K} . If the series \sum u_{n} and \sum v_{n} converge, then the series \sum(u_{n}+\lambda v_{n}) also converges, and we have \sum(u_{n}+\lambda v_{n})=\sum u_{n}+\lambda\sum v_{n}. Let us apply this proposition to prove the convergence of the series \zeta_{M}(s) . Convergence of the Series \sum_{n=2}^{\infty}\frac{1}{x_{n}(s)} We know that x_{n}(s)\geq\left\lfloor\frac{n^{s}}{4}\right\rfloor+1 , a condition that was analytically imposed in the introduction. By definition, for any real number a>0 , we have \left\lfloor a\right\rfloor+1>a . Therefore, x_{n}(s)>\frac{n^{s}}{4}\quad\Rightarrow\quad\frac{1}{x_{n}(s)}<\frac{4}{n^{s}}. (1) Since the general term \frac{4}{n^{s}} defines a Riemann-type series that converges for s>1 , it follows by the comparison theorem that the series \sum_{n=2}^{\infty}\frac{1}{x(s)} also converges. Convergence of the Series \sum_{n=2}^{\infty}\left(\frac{1}{y_{n}(s)}+\frac{1}{z_{n}(s)}\right) We observe that: \left|\frac{1}{y_{n}(s)}+\frac{1}{z_{n}(s)}\right|=\left|\frac{1}{t(4x_{n}(s)-n^{s})-q_{n}(s)}+\frac{1}{t(4x_{n}(s)-n^{s})+q_{n}(s)}\right|. By the standard identity of conjugate reciprocals, this yields: \left|\frac{1}{y_{n}(s)}+\frac{1}{z_{n}(s)}\right|=\left|\frac{2t(4x_{n}(s)-n^{s})}{t^{2}(4x_{n}(s)-n^{s})^{2}-q_{n}(s)^{2}}\right|. But since t^{2}(4x_{n}(s)-n^{s})^{2}-q_{n}(s)^{2}=2tn^{s}x_{n}(s), we have: \left|\frac{1}{y_{n}(s)}+\frac{1}{z_{n}(s)}\right|=\frac{|4x_{n}(s)-n^{s}|}{n^{s}x_{n}(s)}. Now, by the triangle inequality and the positivity of x_{n}(s) and n^{s} , we get: |4x_{n}(s)-n^{s}|\leq 4x_{n}(s)+n^{s}\quad\Rightarrow\quad\frac{|4x_{n}(s)-n^{s}|}{n^{s}x_{n}(s)}\leq\frac{4}{n^{s}}+\frac{1}{x_{n}(s)}. Using inequality ( 1 ), we deduce: \left|\frac{1}{y_{n}(s)}+\frac{1}{z_{n}(s)}\right|<\frac{8}{n^{s}}. Thus, the series \sum_{n=2}^{\infty}\left(\frac{1}{y_{n}(s)}+\frac{1}{z_{n}(s)}\right) converges (absolutely) by comparison with a convergent Riemann series (for s>1 ). Conclusion By the linearity of convergent series, we conclude that the series \zeta_{M}(s)=\sum_{n=2}^{\infty}\left(\frac{1}{x_{n}(s)}+\frac{1}{y_{n}(s)}+\frac{1}{z_{n}(s)}\right) also converges. Therefore, we have the identity: \zeta_{M}(s)=\sum_{n=2}^{\infty}\frac{4}{n^{s}}=4(\zeta(s)-1). (2) Why is this relation interesting? It expresses the discrete Riemann zeta function (which lies at the heart of mathematics) in terms of the series \zeta_{M} , under Mballa’s Conjecture, which itself implies the Erdős–Straus Conjecture. Therefore, the more information we have about Mballa’s Conjecture (in particular, the structure of q_{n}(s) , x_{n}(s) , etc.), the more insight we gain about the Erdős–Straus Conjecture. Most importantly, this also provides a new angle for understanding the values of the zeta function at odd integers, which remain one of the major mysteries in mathematics. Indeed, as stated in ( 2 ), we have: 4(\zeta(2k+1)-1)=\zeta_{M}(2k+1),\quad\text{for }k\in\mathbb{N}^{*}. 5 Numerical Evidence Remark on Numerical Simulations Although numerical simulations can be useful to illustrate the accuracy and convergence behavior of the identity \zeta_{M}(s)=4(\zeta(s)-1), they are not strictly necessary in our case. This identity follows rigorously from the pointwise equality \frac{1}{x_{n}(s)}+\frac{1}{y_{n}(s)}+\frac{1}{z_{n}(s)}=\frac{4}{n^{s}}, which holds for all n\geq 2 under Mballa’s Conjecture. Given that both series converge, the equality of the infinite sums is ensured by classical results on term-by-term summation. Nevertheless, for the interested reader, we provide in the appendix a Python script that numerically verifies this identity up to any chosen value of n , for fixed s\in\mathbb{N}^{*} . References [1] P. U. Mballa, Partial Resolution of the Erdős–Straus, Sierpiński, and Generalized Erdős–Straus Conjectures Using New Analytical Formulas , arXiv:2502.20935 [math.NT], 2025. Available at: http://arxiv.org/abs/2502.20935 [2] P. Erdős, On a Diophantine Equation , Mat. Lapok, vol. 1, pp. 192–210, 1950. [3] J. Derbyshire, The Prime Obsession: Bernhard Riemann and the Greatest Unsolved Problem in Mathematics , 1st ed., Walker & Company, 2003. Appendix: Numerical Verification Code The following Python script allows the reader to verify the identity \zeta_{M}(s)=4(\zeta(s)-1) numerically up to any chosen value N_{\text{max}} , for fixed s\in\mathbb{N}^{*} , s>1 . For example, for s=3 and N_{\text{max}}=150 , the program returns: \zeta(3)-1\approx 0.20203482859170052 Approach via series: 0.20203482859170052 Absolute error: 0.0 This shows a perfect numerical match between both sides of the identity, confirming the analytic equality with machine precision. Below is the full Python script used for the computation: ⬇ 1 import numpy as np 2 3 def find_triplet_corrected ( n , s =2, a =4): 4 "" " 5 Find (x, y, z) in (N^*)^3 such that: 6 4 / n^s = 1/x + 1/y + 1/z, 7 using: 8 y = t(4x - n^s) - q, 9 z = t(4x - n^s) + q, 10 with q^2 = t^2(4x - n^s)^2 - 2tn^s x 11 " "" 12 n_s = n ** s 13 for x in range ( n_s // a + 1, 300 * n_s ): 14 t_min = max (1,

Read the complete text