Systems with discrete singular ϕ \phi -Laplacian and maximal monotone boundary conditions
Andreea GRUIE, Petru JEBELEAN, Călin ŞERBAN · 2026
Read Systems with discrete singular ϕ \phi -Laplacian and maximal monotone boundary conditions by Andreea GRUIE, Petru JEBELEAN, Călin ŞERBAN (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-2604-01998.txt.
Introduction
Gruie, Jebelean and Serban ask when a discrete singular phi Laplacian system has solutions under a broad two point boundary rule. When the added term has no potential form, they prove that solutions exist by a priori estimates. When the term has gradient form, they use an energy method for convex, lower semicontinuous perturbations, finding solutions as low points or saddle points of the energy. The work thus covers both the non potential case and the gradient case in one frame.
Catalog record
| CREATOR | Andreea GRUIE, Petru JEBELEAN, Călin ŞERBAN |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 79518 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2604.01998
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Systems with discrete singular ϕ \phi -Laplacian and maximal monotone boundary conditions Andreea GRUIE, Petru JEBELEAN, Călin ŞERBAN Abstract We are concerned with solvability of nonlinear systems involving a discrete singular ϕ \phi -Laplacian operator of type u ↦ Δ [ ϕ ( Δ u ( n − 1 ) ) ] ( n ∈ { 1 , … , T } ) , u\mapsto\Delta\left[\phi(\Delta u(n-1))\right]\qquad(n\in\{1,\dots,T\}), associated with a general two point boundary condition having the form ( ϕ ( Δ u ( 0 ) ) , − ϕ ( Δ u ( T ) ) ) ∈ γ ( u ( 0 ) , u ( T + 1 ) ) , \left(\phi(\Delta u(0)),-\phi(\Delta u(T))\right)\in\gamma(u(0),u(T+1)), where γ : ℝ N × ℝ N → 2 ℝ N × ℝ N \gamma:\mathbb{R}^{N}\times\mathbb{R}^{N}\to 2^{\mathbb{R}^{N}\times\mathbb{R}^{N}} is a maximal monotone operator with 0 ℝ N × ℝ N ∈ γ ( 0 ℝ N × ℝ N ) 0_{\mathbb{R}^{N}\times\mathbb{R}^{N}}\in\gamma(0_{\mathbb{R}^{N}\times\mathbb{R}^{N}}) . The mapping ϕ \phi is a potential homeomorphism from an open ball of radius a a centered at the origin B a ⊂ ℝ N B_{a}\subset\mathbb{R}^{N} onto ℝ N \mathbb{R}^{N} and Δ \Delta stands for the usual forward difference operator. When the perturbing nonlinearity in the system has not a potential structure we obtain existence of solutions by a priori estimates. Also, when the nonlinearity is of gradient type and γ \gamma is a subdifferential, we provide a variational approach of the system in the frame of critical point theory for convex, lower semicontinuous perturbations of C 1 C^{1} -functionals. Then we derive the existence of solutions either as minimizers or saddle points of the corresponding energy functional. Mathematics Subject Classification: 39A27, 39A12, 39A70, 47J30. Keywords and phrases: discrete singular \phi -Laplacian, maximal monotone operator, subdifferential, a priori estimates, fixed point, critical point, Palais-Smale condition, saddle-point solution. 1 Introduction In this paper we first deal with the unique solvability of the discrete system -\Delta\left[\phi(\Delta u(n-1))\right]+u(n)=h(n)\quad(n\in\mathbb{Z}[1,T]), (1.1) subject to a general two point boundary condition of type \left(\phi(\Delta u(0)),-\phi(\Delta u(T))\right)\in\gamma(u(0),u(T+1)). (1.2) Here and hereafter, for p,q\in\mathbb{Z} with p<q , the notation \mathbb{Z}[p,q] stands for the discrete interval \{p,p+1,\ldots,q\} and \Delta u(n-1):=u(n)-u(n-1) is the usual forward difference operator. We denote by B_{a} the open ball in \mathbb{R}^{N} – endowed with the Euclidean norm |\cdot| , centered in 0_{\mathbb{R}^{N}} of radius a>0 and we assume the following hypotheses on the data entering in problem ( 1.1 )-( 1.2 ): \begin{array}[]{cl}(H_{\Phi})&\phi\,\,\mbox{\sl is a homeomorphism from }B_{a}\,\,\mbox{\sl onto }\mathbb{R}^{N},\,\,\mbox{\sl such that }\phi(0_{\mathbb{R}^{N}})=0_{\mathbb{R}^{N}},\,\phi=\nabla\Phi,\,\,\mbox{\sl with }\Phi:\overline{B}_{a}\to\mathbb{R}\ \mbox{\sl of}\\ &\mbox{\sl class }C^{1}\mbox{\sl on }B_{a},\,\mbox{\sl continuous, strictly convex on }\overline{B}_{a}\mbox{ and }\Phi(0_{\mathbb{R}^{N}})=0;\\ (H_{\gamma})&\gamma:\mathbb{R}^{N}\times\mathbb{R}^{N}\to 2^{\mathbb{R}^{N}\times\mathbb{R}^{N}}\mbox{\sl is maximal monotone with }0_{\mathbb{R}^{N}\times\mathbb{R}^{N}}\in\gamma(0_{\mathbb{R}^{N}\times\mathbb{R}^{N}});\\ (H_{h})&h:\mathbb{Z}[1,T]\to\mathbb{R}^{N}\mbox{\sl is a given discrete function}.\end{array} The inner product \langle\cdot|\cdot\rangle engendering the norm |\cdot| is considered on \mathbb{R}^{N} and the product space \mathbb{R}^{N}\times\mathbb{R}^{N} is endowed with its usual inner product, denoted by \langle\!\langle\cdot|\cdot\rangle\!\rangle . The monotonicity assumption in ( H_{\gamma} ) is understood in the sense of \langle\!\langle\cdot|\cdot\rangle\!\rangle and the notation \|\cdot\| will stand for the norm corresponding to \langle\!\langle\cdot|\cdot\rangle\!\rangle . By a solution of ( 1.1 )-( 1.2 ) we mean a function u:\mathbb{Z}[0,T+1]\to\mathbb{R}^{N} with \|\Delta u\|_{\infty}:=\displaystyle\max_{n\in\mathbb{Z}[1,T+1]}|\Delta u(n-1)|<a\,\mbox{ and }\,(u(0),u(T+1))\in D(\gamma), which satisfies ( 1.1 ) and ( 1.2 ). It is worth to notice that, setting D_{\sigma}:=\{(x,y)\in\mathbb{R}^{N}\times\mathbb{R}^{N}:|x-y|<\sigma\}\quad(\sigma>0), (1.3) if u is a solution of ( 1.1 )-( 1.2 ) then, since |u(T+1)-u(0)|\leq\sum_{n=1}^{T+1}\left|\Delta u(n-1)\right|<(T+1)\,a, actually one has that (u(0),u(T+1))\in D(\gamma)\cap D_{(T+1)\,a}. (1.4) The existence and uniqueness result which we obtain for problem ( 1.1 )-( 1.2 ) (see Theorem 2.1 ) will enable us to discuss the solvability of a general system having the form -\Delta\left[\phi(\Delta u(n-1))\right]=f(n,u(0),\dots,u(T+1))\quad(n\in\mathbb{Z}[1,T]), (1.5) associated with the boundary condition ( 1.2 ), under hypotheses ( H_{\Phi} ), ( H_{\gamma} ) and, instead of ( H_{h} ), (H_{f})\;\;\;f:\mathbb{Z}[1,T]\times(\mathbb{R}^{N})^{T+2}\to\mathbb{R}^{N}\mbox{\sl is continuous.} Since the solutions in the cases of the systems ( 1.1 )-( 1.2 ) and ( 1.5 )-( 1.2 ) are obtained by monotonicity and topological arguments, the question of variational solutions naturally arises. For this, however, it is necessary that the problem to have a potential structure. We will address a situation when such a structure occurs in the case of the system -\Delta\left[\phi(\Delta u(n-1))\right]=\nabla_{u}F(n,u(n))\quad(n\in\mathbb{Z}[1,T]) (1.6) associated with the boundary condition \left(\phi(\Delta u(0)),-\phi(\Delta u(T))\right)\in\partial j(u(0),u(T+1)), (1.7) under hypothesis ( H_{\Phi} ), together with \begin{array}[]{clrc}(H_{F})&F\!=\!F(n,u)\!:\!\mathbb{Z}[1,T]\times\mathbb{R}^{N}\to\mathbb{R}\,\,\mbox{\sl is continuous, }\nabla_{u}F\,\,\mbox{\sl exists and is continuous on the set}\;\mathbb{Z}[1,T]\times\mathbb{R}^{N}\\ &\mbox{\sl and }F(\,\cdot\,,0_{\mathbb{R}^{N}})=0;\\ (H_{j})&j:\mathbb{R}^{N}\times\mathbb{R}^{N}\rightarrow(-\infty,+\infty]\,\,\mbox{\sl is proper, convex, lower semicontinuous, }j(0_{\mathbb{R}^{N}\times\mathbb{R}^{N}})=0\;\mbox{\sl and }\\ &0_{\mathbb{R}^{N}\times\mathbb{R}^{N}}\in\partial j(0_{\mathbb{R}^{N}\times\mathbb{R}^{N}}),\end{array} which respectively replace ( H_{f} ) and ( H_{\gamma} ). By \partial j in ( 1.7 ) we have denoted the subdifferential of j in the sense of convex analysis [ 24 ] . It is useful to recall that if K\subset\mathbb{R}^{N}\times\mathbb{R}^{N} is a nonempty, closed and convex set, then denoting by I_{K} the indicator function of K and by N_{K} the normal cone to K , one has that \partial I_{K}(z)=N_{K}(z) if z\in K and \partial I_{K}(z)=\emptyset if z\notin K (see p. 215-216 in [ 24 ] ). In recent years, much attention has been paid to problems with singular difference operators of type \Delta\left[\phi(\Delta u(n-1))\right] , perturbed with terms having various structures and associated with classical boundary conditions such as Dirichlet, Neumann, Neumann-Steklov, periodic, subharmonic, etc. (see [ 2 , 5 , 8 , 9 , 10 , 11 , 17 , 18 , 19 , 21 ] and the references therein). Such problems are discrete variants of the analogous differential ones and the prototype of \Delta\left[\phi(\Delta u(n-1))\right] is the classical discrete relativistic operator u\mapsto\mathcal{R}(u):=\Delta\left[\frac{\Delta u(n-1)}{\sqrt{1-|\Delta u(n-1)|^{2}}}\right]\qquad(n\in\mathbb{Z}[1,T]), corresponding to \phi(y)={y}/{\sqrt{1-|y|^{2}}} ( y\in B_{1} ). As emphasized in [ 21 ] , the operator \mathcal{R} may be seen as a discretization of the acceleration in special relativity [ 6 , 16 ] . Results on this topic focus on the cases when the unknown function is with either scalar values (most of them) [ 5 , 9 , 10 , 17 , 18 , 19 ] or, as in our case, is \mathbb{R}^{N} -valued [ 2 , 8 , 11 , 21 ] . It is worth to notice that scalar unknown function ( N=1 ) corresponds to the discrete version of a scalar differential equation, while the more general case of \mathbb{R}^{N} -valued unknown function fits with discrete form of a differential system. Approaches to this topic are varied and make use of tools such as fixed point, topological degree, lower and upper solutions, critical point, as well as various combinations of these. It should be emphasized that hypothesis ( H_{\phi} ) – also employed in [ 11 , 21 ] , allows to consider general discrete singular \phi -Laplacians, among which we note the discrete p -relativistic operator u\mapsto\mathcal{R}_{p}(u):=\Delta\left[\frac{|\Delta u(n-1)|^{p-2}\Delta u(n-1)}{(1-|\Delta u(n-1)|^{p})^{1-1/p}}\right]\qquad(n\in\mathbb{Z}[1,T],\,p>1), which is the discrete variant of the differential p -relativistic operator [ 13 ] , or its double phase variant u\mapsto\mathcal{R}_{p}(u)+\mathcal{R}_{q}(u)\quad(p,q>1,\,p\neq q). Observe that \mathcal{R}_{p} is engendered by \phi(y)={|y|^{p-2}y}/{(1-|y|^{p})^{1-1/p}} ( y\in B_{1} ), has the potential \Phi(y)=1-(1-|y|^{p})^{1/p} ( y\in\overline{B}_{1} ) and yields \mathcal{R} for p=2 . Concerning the general boundary condition ( 1.2 ), first of all is worth to notice that it covers the classical ones. Thus, the Dirichlet, Neumann and mixed Dirichlet-Neumann homogeneous boundary conditions \quad u(0)=0_{\mathbb{R}^{N}}=u(T+1), (1.8) \phi(\Delta