Fractional Lane-Emden Hamiltonian systems
Ignacio Ceresa Dussel, Julián Fernández Bonder, Nicolas Saintier and Ariel Salort · 2025
Read Fractional Lane-Emden Hamiltonian systems by Ignacio Ceresa Dussel, Julián Fernández Bonder, Nicolas Saintier and Ariel Salort (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-2501-11523.txt.
Introduction
Ignacio Ceresa Dussel, Julián Fernández Bonder, Nicolas Saintier, and Ariel Salort wrote this math paper in two thousand twenty-five. It proves that a paired system of fractional Lane-Emden equations, of Hamiltonian type, has solutions.
Each equation uses a fractional Laplacian, a nonlocal spread of the classic one, and each unknown feeds the other's rate through a shared potential. Both unknowns must vanish outside a fixed region. The proof leans on a flexible method that traces back to work by De Figueiredo and Felmer. It joins fractional Sobolev spaces with tools for self-adjoint operators. The same path can serve other nonlocal operators, the authors note.
Catalog record
| CREATOR | Ignacio Ceresa Dussel, Julián Fernández Bonder, Nicolas Saintier and Ariel Salort |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 41 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2501-11523
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Fractional Lane-Emden Hamiltonian systems Ignacio Ceresa Dussel, Julián Fernández Bonder, Nicolas Saintier and Ariel Salort Abstract . In this work, our interest lies in proving the existence of solutions to the following Fractional Lane-Emden Hamiltonian system: { ( − Δ ) s u = H v ( x , u , v ) in Ω , ( − Δ ) s v = H u ( x , u , v ) in Ω , u = v = 0 in ℝ n ∖ Ω . \begin{cases}(-\Delta)^{s}u=H_{v}(x,u,v)&\text{in }\Omega,\\ (-\Delta)^{s}v=H_{u}(x,u,v)&\text{in }\Omega,\\ u=v=0&\text{in }\mathbb{R}^{n}\setminus\Omega.\end{cases} The method, that can be traced back to the work of De Figueiredo and Felmer [ 6 ] , is flexible enough to deal with more general nonlocal operators and make use of a combination of fractional order Sobolev spaces together with functional calculus for self-adjoint operators. 2020 Mathematics Subject Classification 35J50, 35J47, 35D30, 35R11 1. Introduction Given a bounded and smooth domain \Omega\subset\mathbb{R}^{n} , n\geq 3 and suitable functions f and g , existence and non-existence of solutions to Hamiltonian systems of the form (1.1) \displaystyle\begin{cases}-\Delta u=f(v)&\text{ in }\Omega\\ -\Delta v=g(u)&\text{ in }\Omega\\ u=v=0&\text{ on }\partial\Omega,\end{cases} was studied in [ 5 , 6 , 10 , 14 ] . These articles provide for a suitable framework to deal with this kind of systems with strongly indefinite variational structure. The prototypical case of ( 1.1 ) was studied in [ 5 ] and it is given when f(v)=v^{p-1} and g(u)=u^{q-1} , where u and v are positive functions: (1.2) \displaystyle\begin{cases}-\Delta u=v^{p-1}&\text{ in }\Omega\\ -\Delta v=u^{q-1}&\text{ in }\Omega\\ u=v=0&\text{ on }\partial\Omega.\end{cases} The authors proved existence of at least positive solutions of this system when (1.3) \displaystyle\frac{n-2}{n}<\frac{1}{p}+\frac{1}{q}<1. On the contrary, when this condition is not fulfilled, it was proved in [ 14 ] that ( 1.2 ) does not admit positive solutions if the domain \Omega is smooth and star-shaped. Observe that this problem can be thought of as a natural extension of the Lane-Emden equation for u\geq 0 : (1.4) -\Delta u=u^{p-1}\text{ in }\Omega,\quad u=0\text{ in }\partial\Omega, which (among many variations) was widely studied and it is well known that its solvability is strongly related with the Sobolev inequality. Indeed it follows from the classical Pohozaev identity that ( 1.4 ) admits a positive solution if and only if p<2^{*}=\frac{2n}{n-2} . Solutions to ( 1.1 ) are obtained as critical points of the Lagrangian (1.5) \displaystyle\mathcal{L}(u,v)=\int_{\Omega}\nabla u\cdot\nabla v\,dx-\int_{\Omega}F(v)\,dx-\int_{\Omega}G(u)\,dx where F and G are primitives of f and g . In contrast with gradient systems, the functional \mathcal{A}(u,v)=\int_{\Omega}\nabla u\cdot\nabla v\,dx , i.e. the quadratic part of \mathcal{L} , is strongly indefinite, in the sense that (0,0) is a saddle point for \mathcal{A} with infinite Morse index. Therefore any decomposition of the solution space into two subspace H_{1} and H_{2} such that \mathcal{A} restricted to H_{1} has a minimum in (0,0) , and \mathcal{A} restricted to H_{2} has a maximum in (0,0) , imply that H_{1} and H_{2} are infinite dimensional. Hence neither Ambrosetti-Rabinowitz’ Mountain Pass Theorem [ 1 ] (which typically assumes a minimum at the origin) nor Rabinowitz’ Saddle Point Theorem [ 15 ] (which requires one of the subspaces H_{1} or H_{2} to be finite dimensional), can be applied in this case. To deal with this kind of problems involving strongly indefinite functionals, some generalizations of the mountain pass and saddle point theorems were developed. We mention for instance the abstract theorems in [ 2 ] , [ 9 ] and [ 8 ] . By means of the topological min-max approach developed in the above mentioned articles, existence and regularity results for solutions to ( 1.2 ) were obtained in [ 5 , 6 , 10 , 14 ] . Later, more general Hamiltonian system of the form (1.6) \displaystyle\begin{cases}-\Delta u=H_{v}(x,u,v)&\text{ in }\Omega\\ -\Delta v=H_{u}(x,u,v)&\text{ in }\Omega\\ u=v=0&\text{ on }\partial\Omega\end{cases} were studied in [ 6 ] , where H includes in particular the prototypical Hamiltonian H(x,u,v)=|u|^{p}+|v|^{q} corresponding to ( 1.2 ) with exponents p,q satisfying ( 1.3 ). In this case, solutions were obtained in [ 6 ] by applying a variational approach using a generalized Mountain Pass Theorem stated in [ 8 ] in the context of Hamiltonian systems, which allows to deal with the strongly indefinite structure of the energy functional related to ( 1.6 ). On the other hand, the last years experimented an increasing attention for non-nocal problems due to their many applications to models of diverse branches of science. One of the most studied nonlocal operator is the well-known fractional Laplacian defined for s\in(0,1) by (-\Delta)^{s}u(x)=\text{p.v. }C(n,s)\int_{\mathbb{R}^{n}}\frac{u(x)-u(y)}{|x-y|^{n+2s}}\,dy, where the constant C(n,s) is positive and depends only on n and s . From now on we will omit for ease of notation. A non local variant of the classical Lane-Emden system ( 1.2 ), namely (1.7) \displaystyle\begin{cases}(-\Delta u)^{s}=v^{p-1}&\text{ in }\Omega,\\ (-\Delta v)^{s}=u^{q-1}&\text{ in }\Omega,\\ u=v=0&\text{ in }\mathbb{R}^{n}\backslash\Omega,\end{cases} where \Omega\subset\mathbb{R}^{n} is a bounded smooth domain in \mathbb{R}^{n} and n>2s , was studied in [ 11 ] in the context of positive viscosity solutions. Solutions to ( 1.7 ) are obtained as critical points of Lagrangian (1.8) \mathcal{L}(u,v)=\frac{1}{2}\int_{\mathbb{R}^{n}}\int_{\mathbb{R}^{n}}\frac{(u(x)-u(y))(v(x)-v(y))}{|x-y|^{n+2s}}\,dxdy-\frac{1}{p}\int_{\Omega}v^{p}\,dx-\frac{1}{q}\int_{\Omega}u^{q}\,dx, which, as in the local case ( 1.5 ), turns out to be strongly indefinite. The authors in [ 11 ] proved that when p,q>1 are such that (1.9) \frac{n-2s}{n}<\frac{1}{p}+\frac{1}{q}<1 then ( 1.7 ) admits at least one positive viscosity solution, and no positive viscosity solution is admitted when the previous condition does not hold and the domain is star-shaped. Observe that assumption ( 1.9 ) simplifies to ( 1.3 ) when s=1 . Moreover ( 1.7 ) can be seen as a generalization of the fractional Lane-Emden problem (-\Delta)^{s}u=u^{p-1}\text{ in }\Omega,\quad u=0\text{ in }\mathbb{R}^{n}\setminus\Omega which by standard variational methods admits a positive energy weak solution if 1<p<2n/(n-2s) , ( p\neq 2 ). In order to prove existence of positive solutions to ( 1.7 ), the idea in [ 11 ] is to observe that v=((-\Delta u)^{s})^{\frac{1}{p-1}} . So, plugging this expression into the second equation in ( 1.7 ), reduces the problem to find critical points of a functional depending only on u instead of the functional ( 1.8 ). This method however does not seem to be adaptable to treat more general Lane-Emden system with Hamiltonian H . Inspired by [ 6 ] we consider in this work the following nonlocal version of ( 1.6 ) (1.10) \begin{cases}(-\Delta)^{s}u=H_{v}(x,u,v)&\text{in }\Omega,\\ (-\Delta)^{s}v=H_{u}(x,u,v)&\text{in }\Omega,\\ u=v=0&\text{in }\mathbb{R}^{n}\setminus\Omega,\end{cases} where \Omega\subset\mathbb{R}^{n} is an open bounded set with smooth boundary and the so-called Hamiltonian function H\colon\Omega\times\mathbb{R}^{2}\to\mathbb{R} is of class C^{1} and satisfies some growths assumptions stated below. We remark that Hamiltonian systems of the form ( 1.10 ) involving fractional Laplacians of different order, let us say, s,t\in(0,1) , lead to non-variational related energy functionals. As far as we know, the techniques introduced in the literature do not seem to be applicable, so we restrict ourselves to the case of two fractional Laplacians with the same order s . The main goal of this article will be to obtain solutions to ( 1.10 ). We want to emphasize that our approach strongly relies on the ideas of [ 6 ] . More precisely, solutions of ( 1.10 ) will be found as critical points of