Constructive Proofs of the Erdos-Straus Conjecture for Prime Numbers with P congruent to 1 modulo 4
E. Dyachenko
TL;DR
This work tackles the Erdos–Straus conjecture for primes $P\equiv1\pmod4$ by introducing two constructive approaches, ED1 (a nonlinear in $P$ parameterization via a factorization identity) and ED2 (a linear-in-$P$ framework yielding an affine lattice in $(\delta,b,c)$). It proves a constructive existence theorem: for every such prime, there exists a representation $4/P=1/A+1/(bP)+1/C$ with $C=cP$ and $(\delta,b,c)$ explicit from ED2, and it develops convolution/anticonvolution to connect the two methods. The paper establishes affine-lattice structure, density estimates, and convergence results for ED2, and validates the theory with large-scale computational verification, including examples up to several thousands in $P$. Overall, the work provides a rigorous, computation-backed pathway to a constructive ES decomposition in the $P\equiv1\pmod4$ case and outlines a program toward a canonical, unified parameterization.
Abstract
The Erdos-Straus conjecture (ESC) concerns the representation of the fraction 4/P, where P is a prime number, as a sum of three positive unit fractions. The focus here is on the case when P is congruent to 1 modulo 4. Two constructive approaches are proposed. Method ED1 is based on a factorization identity and leads to a nonlinear parameterization in P, which requires divisor enumeration and local filtering. Method ED2 yields a linear system in P for the parameters (delta, b, c), describing the solution set as an affine lattice of finite index in Z^3. The central result states that for every prime P congruent to 1 modulo 4 there exists a representation: 4/P = 1/A + 1/(bP) + 1/(cP), where the triple (delta, b, c) in N^3 is constructed explicitly by method ED2. In addition, algorithms for transforming solutions (convolution and anti-convolution) are introduced, and large-scale computational verification confirms the correctness and efficiency of the proposed methods.
