Existence Result for Difference Equations on Non-Uniform Grids via Upper and Lower Solution Method
Shalmali Bandyopadhyay, Kimsear Lor · 2025
Read Existence Result for Difference Equations on Non-Uniform Grids via Upper and Lower Solution Method by Shalmali Bandyopadhyay, Kimsear Lor (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-2508-04706.txt.
Introduction
Shalmali Bandyopadhyay and Kimsear Lor prove that certain discrete equations on unevenly spaced grids always have a solution trapped between two bounding functions. These difference equations model processes sampled at irregular times, and the boundary conditions here are mixed and non-homogeneous, which complicates the math. Earlier results covered only simpler boundary setups.
Their strategy separates the boundary effects from the equation's core structure. That split lets them apply a classic fixed point theorem inside an affine function space. The result guarantees a solution lying between a chosen lower and upper function. The work extends the upper and lower solution method to a broader class of discrete boundary value problems.
Catalog record
| CREATOR | Shalmali Bandyopadhyay, Kimsear Lor |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 40 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2508-04706
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Existence Result for Difference Equations on Non-Uniform Grids via Upper and Lower Solution Method Shalmali Bandyopadhyay, Kimsear Lor Abstract . This paper establishes an existence theory for discrete second-order boundary value problems on non-uniform time grids using the upper and lower solution method. We consider difference equations of the form u Δ Δ ( t i − 1 ) + f ( t i , u ( t i ) , u Δ ( t i − 1 ) ) = 0 u^{\Delta\Delta}(t_{i-1})+f(t_{i},u(t_{i}),u^{\Delta}(t_{i-1}))=0 on a non-uniform time grid t 0 , t 1 , … , t n + 2 {t_{0},t_{1},\ldots,t_{n+2}} with mixed boundary conditions u Δ ( t 0 ) = 0 u^{\Delta}(t_{0})=0 and u ( t n + 2 ) = g ( t n + 2 ) u(t_{n+2})=g(t_{n+2}) . This extends previous work on homogeneous boundary conditions to the non-homogeneous case, requiring a sophisticated functional analytic framework to handle the resulting affine function spaces. Our approach employs a decomposition strategy that separates boundary effects from the differential structure, enabling the application of Brouwer’s Fixed Point Theorem to establish existence with solutions bounded between upper and lower functions. Keywords: Discrete time scale, Difference equations, Non-uniform step size, Boundary value problem, Non-homogeneous mixed boundary, Upper and lower solution method Mathematics Subject Classification (2020): 34B16, 34B18, 34B40, 39A10 1. Introduction Difference equations arise naturally in numerous scientific and engineering applications where discrete time modeling is essential or preferred over continuous formulations. In computational finance discrete-time models are fundamental for option pricing, where finite difference methods are used (e.g. see [ 13 ] ) to price options by approximating continuous-time differential equations with discrete-time difference equations, building on foundational work from the 1970s. Economic modeling extensively employs [ 7 ] discrete time formulations, as economic behavior is inherently dynamic and much of economic analysis is based on discrete time periods such as months, quarters, or years, reflecting the periodic nature of data collection and decision-making in financial markets and policy analysis. Simultaneously, population dynamics modeling represents a crucial domain where difference equations have a long history of use as discrete time models (see [ 3 ] ), describing typically autonomous discrete time dynamics. Mathematical biology applications continue to expand, where discrete Lotka-Volterra models are more appropriate than continuous ones when populations are of non-overlapping generations, with studies showing that discrete time models governed by difference equations provide more realistic descriptions of population interactions ( [ 4 ] ). Contemporary epidemiological applications have demonstrated the importance of [ 5 , 11 ] discrete SEIR models, where discrete forms are discussed to describe viral disease propagation and find control mechanisms for rapid spread of infectious diseases like COVID-19. Engineering applications particularly benefit from digital signal processing, where difference equations characterize digital filters and provide the mathematical foundation for implementing finite impulse response and infinite impulse response filters in real-time systems (e.g. [ 14 , 2 ] ). Additionally, control theory applications utilize difference equations ( [ 1 ] ) for discrete-time system analysis, where linear constant-coefficient difference equations model linear time-invariant systems and enable the design of digital controllers for various engineering applications. Classical applications in [ 15 ] non-uniform grid discretizations continue to show that careful consideration of truncation errors introduced by variable step sizes is essential for accurate numerical solutions in computational fluid dynamics and partial differential equation solvers. The theoretical foundations are well established in [ 6 , 12 ] comprehensive treatments of discrete chaos and complex systems modeling. In this manuscript we establish existence result of a difference equation on nonuniform grid with mixed boundary condition by employing standard upper and lower solution method, extensively available in the literature together with Brouwer fixed point theory. Prior to stating our problem, we begin by establishing the mathematical framework for non-uniform time grids and associated difference calculus. For further details regarding difference calculus, we refer readers to the book [ 8 ] . Definition 1.1 . Let n\in\mathbb{N} be fixed and a<b be real numbers. A non-uniform time grid is a strictly increasing sequence \mathbb{T}=\{t_{0},t_{1},\ldots,t_{n+2}\} where a=t_{0}<t_{1}<\ldots<t_{n+2}=b . Definition 1.2 . Define the step sizes h_{i}=t_{i+1}-t_{i} for i=0,1,\ldots,n+1 , and assume there exist constants h_{\min},h_{\max}>0 such that 0<h_{\min}\leq h_{i}\leq h_{\max} for all i . Definition 1.3 . For a function u:\mathbb{T}\to\mathbb{R} , the delta derivative at t_{i} is defined by u^{\Delta}(t_{i})=\frac{u(t_{i+1})-u(t_{i})}{t_{i+1}-t_{i}}=\frac{u(t_{i+1})-u(t_{i})}{h_{i}} for i=0,1,\ldots,n+1 . Definition 1.4 . The second delta derivative is defined as u^{\Delta\Delta}(t_{i})=\frac{u^{\Delta}(t_{i+1})-u^{\Delta}(t_{i})}{t_{i+1}-t_{i}}=\frac{u^{\Delta}(t_{i+1})-u^{\Delta}(t_{i})}{h_{i}} for i=0,1,\ldots,n . Having introduced necessary definitions, we now consider the following boundary value problem. \begin{cases}u^{\Delta\Delta}(t_{i-1})+{f}(t_{i},u(t_{i}),u^{\Delta}(t_{i-1}))=0,&i\in\{1,2,\ldots,n+1\}\\ u^{\Delta}(t_{0})=0\\ u(t_{n+2})=g(t_{n+2})\end{cases} (1.1) where \{t_{0},t_{1},\ldots,t_{n+2}\} is a non-uniform partition of an interval [a,b] with a=t_{0}<t_{1}<\ldots<t_{n+2}=b . Moreover, we assume (A1) f(t_{i},\cdot,\cdot) is continuous on \{t_{1},\ldots,t_{n+1}\} , right continuous at t_{0} and left continuous at t_{n+2} in the discrete sense; f(\cdot,x,y) is continuous on \mathbb{R}^{2} . (A2) {f} is non-increasing in its third variable (A3) g:\mathbb{T}\to\mathbb{R}^{+} is continuous (in the discrete sense) In [ 10 ] , authors studied the following problem, \begin{cases}u^{\Delta\Delta}(t_{i-1})+{f}(t_{i},u(t_{i}),u^{\Delta}(t_{i-1}))=0,&i\in\{1,2,\ldots,n+1\}\\ u^{\Delta}(t_{0})=0\\ u(t_{n+2})=0.\end{cases} Later, in [ 9 ] , authors extended the above problem for n^{\text{th}} order derivative. In our case, the second boundary condition allows for non-homogeneous boundary values. The main challenge we faced was defining the underlying function space to accommodate this non-homogeneous structure. This problem requires a more sophisticated functional analytic approach compared to the standard homogeneous case. While homogeneous boundary conditions naturally define linear function spaces that are well-suited for fixed point theory applications, non-homogeneous conditions create affine spaces that lack the linear structure necessary for classical fixed point arguments. To overcome this challenge, we employ a careful decomposition strategy that separates the boundary effects from the core differential structure, allowing the problem to be reformulated within an appropriate linear framework while preserving the original boundary requirements (see Section 2 ). Prior to stating our theorem, we will provide definitions of solution, upper solution and lower solution. Definition 1.5 . A function u:\mathbb{T}\to\mathbb{R} is a solution of ( 1.1 ) if it satisfies both the difference equation and the boundary conditions. Definition 1.6 . A function \alpha:\mathbb{T}\to\mathbb{R} is called a lower solution of ( 1.1 ) if: \begin{cases}\alpha^{\Delta\Delta}(t_{i-1})+{f}(t_{i},\alpha(t_{i}),\alpha^{\Delta}(t_{i-1}))\geq 0,&i\in\{1,\ldots,n+1\}\\ \alpha^{\Delta}(t_{0})\geq 0\\ \alpha(t_{n+2})\leq g(t_{n+2})\end{cases} (1.2) Definition 1.7 . A function \beta:\mathbb{T}\to\mathbb{R} is called an upper solution of ( 1.1 ) if: \begin{cases}\beta^{\Delta\Delta}(t_{i-1})+{f}(t_{i},\beta(t_{i}),\beta^{\Delta}(t_{i-1}))\leq 0,&i\in\{1,\ldots,n+1\}\\ \beta^{\Delta}(t_{0})\leq 0\\ \beta(t_{n+2})\geq g(t_{n+2})\end{cases} (1.3) Theorem 1.8 . Let \alpha and \beta be lower and upper solutions of ( 1.1 ), respectively, with \alpha(t_{i})\leq\beta(t_{i}) for all i\in\{0,1,\ldots,n+1\} . Assume conditions (A1) - (A3) hold. Then ( 1.1 ) has a solution u satisfying \alpha(t_{i})\leq u(t_{i})\leq\beta(t_{i}),\quad i\in\{0,1,\ldots,n+2\} (1.4) The remainder of this paper is organized as follows. Section 2 establishes the functional analytic framework, including the definition of appropriate function spaces, the solution decomposition theorem that separates boundary conditions from the difference equation, and Brouwer’s Fixed Point Theorem in the discrete setting. Section 3 presents the main existence result through three key steps: construction of an auxiliary problem with modified nonlinearity that handles solutions outside the upper and lower bounds, development of a solution operator whose fixed points correspond to solutions of the auxiliary problem, and application of Brouwer’s theorem to establish existence. Finally, Section 4 completes the proof by verifying that any solution of auxiliary