Existence and multiplicity of solutions for fractional p 1 ( x , ⋅ ) & p 2 ( x , ⋅ ) $p_{1}(x,\cdot )\& p_{2}(x,\cdot )$ -Laplacian Schrödinger-type equations with Robin boundary conditions

In this paper, we study fractional p 1 (x , ⋅) & p 2 (x , ⋅) -Laplacian Schrödinger-type equations for Robin boundary conditions. Under some suitable assumptions, we show that two solutions exist using the mountain pass lemma and Ekeland's variational principle. Then, the existence of infinitely many solutions is derived by applying the fountain theorem and the Krasnoselskii genus theory, respectively. Different from previous results, the topic of this paper is the Robin boundary conditions in R N ∖ Ω ‾ for fractional order p 1 (x , ⋅) & p 2 (x , ⋅) -Laplacian Schrödinger-type equations, including concave-convex nonlinearities, which has not been studied before. In addition, two examples are given to illustrate our results.

Keywords: Schrödinger equations; p1(x,⋅)&p2(x,⋅)\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$p{1}(x,\cdot)\& p{2}(x,\cdot)$\end{document} -Laplacian; Robin boundary conditions; Concave-convex nonlinearities; Krasnoselskii genus theory

Tianqing An, Weichun Bu and Shuai Li contributed equally to this work.

Introduction and the main results

In this paper, we consider fractional $p_1(x,\cdot )\mathrm{\&}{p}_2(x,\cdot )$ -Laplacian Schrödinger-type equations, including concave-convex nonlinearities with nonlocal Robin boundary conditions

1 $\{\begin{array}{c}{\sum }_{i=1}^2[{(-\mathrm{\Delta })}_{p_i(x,\cdot )}^s\phi +V_i(x)|\phi {|}^{{\overline{p}}_i(x)-2}\phi ]\hfill \\ ={\lambda }_1{A}_1(x)|\phi {|}^{r_1(x)-2}\phi +{\lambda }_2{A}_2(x)|\phi {|}^{r_2(x)-2}\phi ,x\in \mathrm{\Omega },\hfill \\ {\sum }_{i=1}^2[{\mathcal{N}}_{s,p_i(x,\cdot )}\phi +\beta (x)|\phi {|}^{{\overline{p}}_i(x)-2}\phi ]={\sum }_{i=1}^2{g}_i(x),x\in {\mathbb{R}}^N\setminus \bar{\mathrm{\Omega }},\hfill \end{array}$

Graph

where $V_i(x)$ ( $x\in \mathrm{\Omega }$ , $i=1,2$ ) is a potential function, $\mathrm{\Omega }\subset {\mathbb{R}}^N$ ( $N\ge 2$ ) is a bounded domain with the Lipschitz boundary ∂Ω, $s\in (0,1)$ , $p_i(x,\cdot ):{\mathbb{R}}^{2N}\to (1,+\mathrm{\infty })$ , ${\bar{p}}_i(x)=p_i(x,x)$ , $r_1(x)$ , $r_2(x)$ are continuous functions, ${\lambda }_1$ , ${\lambda }_2$ are positive constants, $A_1(x)$ , $A_2(x)$ are positive weighted functions, $g_i(x)\ge 0\in L^1({\mathbb{R}}^N\setminus \mathrm{\Omega })$ , $\beta (x)\ge 0\in L^{\mathrm{\infty }}({\mathbb{R}}^N\setminus \mathrm{\Omega })$ ,

${\mathcal{N}}_{s,p_i(x,\cdot )}\phi (x)={\int }_{\mathrm{\Omega }}\frac{|\phi (x)-\phi (y){|}^{p_i(x,y)-2}(\phi (x)-\phi (y))}{|x-y{|}^{N+s{p}_i(x,y)}}dy,x\in {\mathbb{R}}^N\setminus \bar{\mathrm{\Omega }},$

Graph

and

${(-\mathrm{\Delta })}_{p_i(x,\cdot )}^s\phi (x):=P.V.{\int }_{{\mathbb{R}}^N}\frac{|\phi (x)-\phi (y){|}^{p_i(x,y)-2}(\phi (x)-\phi (y))}{|x-y{|}^{N+s{p}_i(x,y)}}dy,x\in \mathrm{\Omega },$

Graph

where $P.V$ . stands for the Cauchy principal value.

Equations (1) arise from general reaction-diffusion equation

2 ${\phi }_t=\mathrm{\nabla }\cdot [A(\phi )\mathrm{\nabla }\phi ]+r(x,\phi ),$

Graph

where $A(\phi )={|\mathrm{\nabla }\phi |}^{p-2}+{|\mathrm{\nabla }\phi |}^{q-2}$ . Problem (2) has applications in biophysics, plasma physics, and chemical reactions. For more details on equation (2), readers are referred to [[1]]. Combining with a ${\mathbb{Z}}_2$ -symmetric version of the mountain pass lemma for even functionals and some adequate variational methods, Mihăilescu [[3]] proved that the equations

3 $\{\begin{array}{c}-div((|\mathrm{\nabla }\phi {|}^{p_1(x)}+|\mathrm{\nabla }\phi {|}^{p_2(x)})\mathrm{\nabla }\phi )=f(x,\phi ),x\in \mathrm{\Omega },\hfill \\ \phi =0,x\in \partial \mathrm{\Omega }\hfill \end{array}$

Graph

have infinitely many weak solutions. In addition, Chung and Toan [[4]] considered a class of fractional Laplacian problems

4 $\{\begin{array}{c}{(\mathrm{\Delta })}_{p_1(x,\cdot )}^s\phi (x)+{(\mathrm{\Delta })}_{p_2(x,\cdot )}^s\phi (x)+|\phi {|}^{q(x)-2}\phi \hfill \\ =\lambda A_1(x)|\phi (x){|}^{r_1(x)-2}\phi (x)-\mu A_2(x)|\phi (x){|}^{r_2(x)-2}\phi (x),x\in \mathrm{\Omega },\hfill \\ \phi (x)=0,x\in \partial \mathrm{\Omega }\hfill \end{array}$

Graph

using variational techniques and Ekeland's variational principle. The authors used the variational techniques to discuss the results of the existence of solutions in fractional cases [[5]–[7]]. In addition, Heidarkhani et al. [[8]–[10]] studied the existence results of variable exponent equations using variational methods and established the critical point theory. Zuo et al. [[11]] investigated the existence and multiplicity of solutions for the $p(x,\cdot )\mathrm{\&}q(x,\cdot )$ fractional Choquard problems with variable order. On a similar issue, a related study was conducted by Biswas et al. For more details, see [[12]].

The classical Schrödinger equation is of the following form:

$ih\frac{\partial }{\partial t}\phi =-\frac{h^2}{2m}{\mathrm{\nabla }}^2\phi +V\phi ,$

Graph

where V, φ denote the potential function and wave function, respectively, and i, h are constants ([[13]]). Recently, Xiang et al. in [[14]] and Bu et al. in [[15]] discussed the fractional Laplace operator Schrödinger equations with variable order and Schrödinger–Kirchhof-type equations, respectively.

The critical local problem involving concave-convex nonlinearities was first studied by Ambrosetti et al. in [[16]]. Subsequently, variational methods were used [[17]] to discuss the following equations:

5 $\{\begin{array}{c}-div(\omega (x)|\mathrm{\nabla }\phi {|}^{p(x)-2}\mathrm{\nabla }\phi )=\lambda a(x)|\phi {|}^{q(x)-2}\phi +\mu b(x)|\phi {|}^{h(x)-2}\phi ,x\in \mathrm{\Omega },\hfill \\ \phi =0,x\in \partial \mathrm{\Omega }\hfill \end{array}$

Graph

with the variable order concave-convex term. For other similar types of equations, see [[18]] and the references therein.

The Robin and Neumann boundary problems are interesting topics [[20]]. Mugnai et al. [[21]] investigated fractional p-Laplacian problems with nonlocal Neumann boundary conditions. Moreover, Deng [[22]] considered the following equations:

6 $\{\begin{array}{c}-{\mathrm{\Delta }}_{p(x)}\phi =\lambda f(x,\phi ),x\in \mathrm{\Omega },\hfill \\ |\mathrm{\nabla }\phi {|}^{p(x)-2}\frac{\partial \phi }{\partial \eta }+\beta (x)|\phi {|}^{p(x)-2}\phi =0,x\in \partial \mathrm{\Omega }.\hfill \end{array}$

Graph

For double-phase problems depending on Robin and Steklov eigenvalues for the p-Laplacian, Manouni et al. [[23]] proved the existence of solutions by variational tools, truncation techniques, and comparison methods. In many papers, the Robin and Neumann boundary problems of fractional equations were studied in different ways; e.g., the Morse theory was used in [[24]], the mountain pass lemma in [[26]], Ekeland's variational principle in [[28]], and the topological degree in [[30]].

To our knowledge, there is no previous work on the problem (1). This paper is devoted to this topic. We obtain new results by applying the mountain pass lemma, Ekeland's variational principle, the fountain theorem, and the Krasnoselskii genus theory. Our problem differs from problems (3), (4), and (5) in that we discuss Robin boundary conditions, and it also differs from problem the (6) in that we consider $p_1(x,\cdot )\mathrm{\&}{p}_2(x,\cdot )$ -Laplacian Schrödinger-type equations with concave-convex nonlinearities.

Before stating the main results, we introduce the basic assumptions.

• $p_i(x,y)$ is a symmetric and continuous function, that is,
• $p_i(x,y)=p_i(y,x),\text{for all}(x,y)\in {\mathbb{R}}^N\times {\mathbb{R}}^N$

Graph

• with

• $1<p_i^{-}:=\underset{(x,y)\in {\mathbb{R}}^N\times {\mathbb{R}}^N}{min}{p}_i(x,y)\le p_i(x,y)\le p_i^{+}:=\underset{(x,y)\in {\mathbb{R}}^N\times {\mathbb{R}}^N}{max}{p}_i(x,y)<+\mathrm{\infty },$

Graph

• and

• $1<p_1^{-}\le p_1^{+}\le p_2^{-}\le p_2^{+}<+\mathrm{\infty },$

Graph

• such that $s{p}_i^{+}<N$ . Let $0<s<1<p(x,\cdot )$ , the fractional critical exponent $p_s^{\ast }(x)$ be defined as $p_s^{\ast }(x)=\frac{Np(x,x)}{N-sp(x,x)}$ and $p(x,\cdot )<p_s^{\ast }(x)$ for all $x\in \bar{\mathrm{\Omega }}$ .
• $\int g_i(x)\phi dx=-\int g_i(x)\phi dx$ .
• $V_i(x)$ is a continuous function, satisfying ${inf}_{x\in \mathrm{\Omega }}{V}_i(x)>V_{i0}>0$ , for all $d_i>0$ , $\mathrm{means}(\{x\in \mathrm{\Omega }:V_i(x)<d_i\})<+\mathrm{\infty }$ .
• $A_1(x)$ and $A_2(x)$ are weighted functions in $C(\bar{\mathrm{\Omega }})$ and satisfy $A_1(x)\in L^{s_1(x)}(\mathrm{\Omega })$ such that $1<s_1(x)\in C(\bar{\mathrm{\Omega }})$ and $1<s_1^{\prime }(x)r_1(x)<p_s^{\ast }(x)$ for all $x\in \bar{\mathrm{\Omega }}$ , $A_2(x)\in L^{s_2(x)}(\mathrm{\Omega })$ such that $1<s_2(x)\in C(\bar{\mathrm{\Omega }})$ and $1<s_2^{\prime }(x)r_2(x)<p_s^{\ast }(x)$ for all $x\in \bar{\mathrm{\Omega }}$ . Here, $s_1^{\prime }(x)$ and $s_2^{\prime }(x)$ are conjugate exponents of the functions $s_1(x)$ and $s_2(x)$ , respectively.

The main results of this paper are as follows:

Theorem 1.1

Assume that assumptions (P), (G), (V), and (H) hold. Equations (1) have two nontrivial weak solutions.

Theorem 1.2

Assume that assumptions (P), (G), (V), and (H) hold. Then, equations (1) have infinitely many nontrivial weak solutions inX.

Theorem 1.3

Assume that assumptions (P), (G), (V), and (H) hold. Then, equations (1) possess infinitely many solutions.

In Sect. 2, we state some basic results of the Lebesgue space $L^{q(x)}(\mathrm{\Omega })$ . In Sect. 3, we introduce the workspaces associated with equations (1). In Sect. 4, we verify the $(PS)$ conditions and prove Theorem 1.1 by the mountain pass lemma and Ekeland's variational principle. In Sect. 5, we prove Theorem 1.2 by applying the fountain theorem. Finally, using the Krasnoselskii genus theory, we give the proof of Theorem 1.3.

Preliminaries

In this section, we recall some basic results of the Lebesgue space $L^{q(x)}(\mathrm{\Omega })$ with a variable exponent. Assume that domain Ω is bounded in ${\mathbb{R}}^N$ with the Lipschitz boundary ∂Ω. Let

$q^{-}=\underset{x\in \bar{\mathrm{\Omega }}}{min}q(x),q^{+}=\underset{x\in \bar{\mathrm{\Omega }}}{max}q(x),$

Graph

where $C_{+}(\bar{\mathrm{\Omega }})=\{q\in C(\bar{\mathrm{\Omega }}):q(x)>1,\text{for all}x\in \bar{\mathrm{\Omega }}\}$ .

The variable exponent Lebesgue space $L^{q(x)}(\mathrm{\Omega })$ , which is defined by

$L^{q(x)}(\mathrm{\Omega })=\{\phi |\phi :\mathrm{\Omega }\to \mathbb{R}\text{is measurable and}{\int }_{\mathrm{\Omega }}|\phi (x){|}^{q(x)}dx<\mathrm{\infty }\},$

Graph

equipped with the Luxemburg norm

${\parallel \phi \parallel }_{q(x)}=inf\{\iota >0:{\int }_{\mathrm{\Omega }}|\frac{\phi {(x)}^{q(x)}}{\iota }|dx\le 1\},$

Graph

where $(L^{q(x)}(\mathrm{\Omega }),{\parallel \cdot \parallel }_{q(x)})$ is a separable, uniformly convex, and reflexive Banach space [[31]].

Let $L^{q^{\prime }(x)}(\mathrm{\Omega })$ be the conjugate space of $L^{q(x)}(\mathrm{\Omega })$ and $1/q(x)+1/q^{\prime }(x)=1$ ( $p(x)$ and $q^{\prime }(x)$ are conjugate indices to each other). For $\phi \in L^{q(x)}(\mathrm{\Omega })$ and $v\in L^{q^{\prime }(x)}(\mathrm{\Omega })$ , the Hölder inequality

7 $|{\int }_{\mathrm{\Omega }}\phi (x)v(x)dx|<(\frac{1}{q^{-}}+\frac{1}{q^{\prime -}}){\parallel \phi \parallel }_{q(x)}{\parallel v\parallel }_{q^{\prime }(x)}\le 2{\parallel \phi \parallel }_{q(x)}{\parallel v\parallel }_{q^{\prime }(x)}$

Graph

holds. If $q_i(x)\in C_{+}(\bar{\mathrm{\Omega }})$ ( $i=1,2,...,\hat{n}$ ) and

$\frac{1}{q_1(x)}+\frac{1}{q_2(x)}+\cdots +\frac{1}{q_{\hat{n}}(x)}=1,$

Graph

for all ${\phi }_i(x)\in L^{q_i(x)}(\mathrm{\Omega })$ , there exists

$|{\int }_{\mathrm{\Omega }}{\phi }_1(x){\phi }_2(x)\cdots {\phi }_{\hat{n}}(x)dx|\le (\frac{1}{q_1^{-}}+\frac{1}{q_2^{-}}+\cdots \frac{1}{q_{\hat{n}}^{-}}){\parallel {\phi }_1\parallel }_{q_1(x)}{\parallel {\phi }_2\parallel }_{q_2(x)}\cdots {\parallel {\phi }_{\hat{n}}\parallel }_{q_{\hat{n}}(x)}.$

Graph

Lemma 2.1

([[32]]) Let ${\rho }_{q(x)}$ be the modular of the $L^{q(x)}(\mathrm{\Omega })$ space, and ${\rho }_{q(x)}:L^{q(x)}(\mathrm{\Omega })\to \mathbb{R}$ defined by ${\rho }_{q(x)}(\phi )={\int }_{\mathrm{\Omega }}{|\phi (x)|}^{q(x)}dx$ . Then, the following properties hold:

• ${\parallel \phi \parallel }_{q(x)}<1(=1,>1)⟺{\rho }_{q(x)}(\phi )<1(=1,>1)$ ;
• ${\parallel \phi \parallel }_{q(x)}>1⟹{\parallel \phi \parallel }_{q(x)}^{q^{-}}\le {\rho }_{q(x)}(\phi )\le {\parallel \phi \parallel }_{q(x)}^{q^{+}}$ ;
• ${\parallel \phi \parallel }_{q(x)}<1⟹{\parallel \phi \parallel }_{q(x)}^{q^{+}}\le {\rho }_{q(x)}(\phi )\le {\parallel \phi \parallel }_{q(x)}^{q^{-}}$ .

Lemma 2.2

([[32]]) If $\phi ,{\phi }_n\in L^{q(x)}(\mathrm{\Omega })$ with $n\in \mathbb{N}$ , then

• ${lim}_{n\to +\mathrm{\infty }}{\parallel {\phi }_n-\phi \parallel }_{q(x)}=0$ ;
• ${lim}_{n\to +\mathrm{\infty }}{\rho }_{q(x)}({\phi }_n-\phi )=0$ ;
• ${\phi }_n(x)\to \phi (x)$ a. e. in Ω and ${lim}_{n\to +\mathrm{\infty }}{\rho }_{q(x)}({\phi }_n)={\rho }_{q(x)}(\phi )$ .

Lemma 2.3

([[33]]) Let $p(x)$ , $q(x)$ be measurable functions such that $p(x)\in L^{\mathrm{\infty }}({\mathbb{R}}^N)$ and $1<p(x)q(x)<\mathrm{\infty }$ , for any $x\in {\mathbb{R}}^N$ . Then, there is

$min\{{\parallel \phi \parallel }_{p(x)q(x)}^{p^{+}},{\parallel \phi \parallel }_{p(x)q(x)}^{p^{-}}\}\le {\parallel |\phi {|}^{p(x)}\parallel }_{q(x)}\le max\{{\parallel \phi \parallel }_{p(x)q(x)}^{p^{+}},{\parallel \phi \parallel }_{p(x)q(x)}^{p^{-}}\}$

Graph

with $\phi \in L^{q(x)}({\mathbb{R}}^N)$ , $\phi \ne 0$ .

The basic properties of functionals and operators

In this section, we state some properties of functionals and operators, and give the definition of weak solutions of equations (1) with Robin boundary conditions. We first introduce the workspaces $(W,{\parallel \cdot \parallel }_W)$ and $(X,{\parallel \cdot \parallel }_X)$ associated with equations (1).

The fractional variable Sobolev space $W:=W^{s,q(x),p(x,\cdot )}(\mathrm{\Omega })$ is given by

$\begin{array}{rl}{W}^{s,q(x),p(x,\cdot )}(\mathrm{\Omega }):=& \{\phi :\mathrm{\Omega }\to \mathbb{R}|\phi \in L^{q(x)}(\mathrm{\Omega }),\\ & {\int }_{\mathrm{\Omega }\times \mathrm{\Omega }}\frac{|\phi (x)-\phi (y){|}^{p(x,y)}}{{\mu }^{p(x,y)}|x-y{|}^{N+sp(x,y)}}dxdy<\mathrm{\infty },\text{for some}\mu >0\}.\end{array}$

Graph

Set

${[\phi ]}_{s,p(x,\cdot )}=inf\{\mu >0:{\int }_{\mathrm{\Omega }\times \mathrm{\Omega }}\frac{|\phi (x)-\phi (y){|}^{p(x,y)}}{{\mu }^{p(x,y)}|x-y{|}^{N+sp(x,y)}}dxdy<1\}$

Graph

as the variable exponent Gagliardo seminorm. W is a Banach space with the norm

${\parallel \phi \parallel }_W={\parallel \phi \parallel }_{Ws,q(x),p(x,\cdot )}={\parallel \phi \parallel }_{q(x)}+{[\phi ]}_{s,p(x,\cdot )}.$

Graph

We take into account three continuous functions $p(x,y):\bar{\mathrm{\Omega }}\times \bar{\mathrm{\Omega }}\to (1,\mathrm{\infty })$ and $r_1(x),r_2(x)\in C_{+}(\bar{\mathrm{\Omega }})$ . From condition (P), we know that

8 $\begin{array}{rl}& p(x,y)=p(y,x),\text{for all}x,y\in \bar{\mathrm{\Omega }};\\ & 1<p^{-}:=minp(x,y)\le p(x,y)\le p^{+}:=maxp(x,y)<\mathrm{\infty };\\ & 1<r_1^{-}:=min{r}_1(x)\le r_1(x)\le r_1^{+}:=max{r}_1(x)<\mathrm{\infty };\\ & 1<r_2^{-}:=min{r}_2(x)\le r_2(x)\le r_2^{+}:=max{r}_2(x)<\mathrm{\infty }.\end{array}$

Graph

Lemma 3.1

([[34]]) Suppose that $\mathrm{\Omega }\subset {\mathbb{R}}^N$ is a bounded open domain and (8) holds. Then, Wis a separable and reflexive space.

Lemma 3.2

([[35]]) Let smooth bounded domain $\mathrm{\Omega }\subset {\mathbb{R}}^N$ , $sp(x,y)<N$ for $(x,y)\in \bar{\mathrm{\Omega }}\times \bar{\mathrm{\Omega }}$ with $s\in (0,1)$ , and $q(x)\ge p(x,x)$ for $x\in \bar{\mathrm{\Omega }}$ . Suppose that continuous function $\hat{h}(x):\bar{\mathrm{\Omega }}\to (1,\mathrm{\infty })$ satisfies

$p_s^{\ast }(x)\ge \hat{h}(x)\ge {\hat{h}}^{-}=\underset{x\in \mathrm{\Omega }}{inf}\hat{h}(x)>1,\mathit{\text{for all}}x\in \bar{\mathrm{\Omega }}.$

Graph

There exists a positive constant $C_0=C_0(N,s,p,\hat{h},\mathrm{\Omega })$ such that for every $\phi \in W$ , it holds that

${\parallel \phi \parallel }_{\hat{h}(x)}\le C_0{\parallel \phi \parallel }_W.$

Graph

Then, the embedding $W\hookrightarrow L^{\hat{h}(x)}$ for all $\hat{h}\in (1,p_s^{\ast })$ is compact.

Lemma 3.3

([[35]]) If $1<s{p}^{-}$ and

$p_{\partial }^{\ast }(x):=\frac{(N-1)p(x,x)}{N-sp(x,x)}>\hat{h}(x),\mathit{\text{in}}\partial \mathrm{\Omega }\cap \{x\in \bar{\mathrm{\Omega }}:N-sp(x,x)>0\}.$

Graph

There exists a positive constant $C_1=C_1(N,s,p,\hat{h},\partial \mathrm{\Omega })$ such that

${\parallel \phi \parallel }_{L^{\hat{h}(x)}(\partial \mathrm{\Omega })}\le C_1{\parallel \phi \parallel }_{s,\bar{p}(x),p(x,y)},\mathit{\text{for all}}\phi \in W^{s,\bar{p}(x),p(x,y)}(\mathrm{\Omega }).$

Graph

Then, the embedding $W^{s,\bar{p}(x),p(x,y)}(\mathrm{\Omega })\hookrightarrow L^{\hat{h}(x)}(\partial \mathrm{\Omega })$ is compact.

Define nonlinear map $\mathcal{L}:W\to W^{\ast }$

9 $\langle \mathcal{L}(\phi ),\psi \rangle ={\int }_{\mathrm{\Omega }}\frac{|\phi (x)-\phi (y){|}^{p(x,y)-2}(\phi (x)-\phi (y))(\psi (x)-\psi (y))}{|x-y{|}^{N+sp(x,y)}}dy,$

Graph

for all $\phi ,\psi \in W$ , $\mathcal{L}$ has the following properties.

Lemma 3.4

([[28]])

• $\mathcal{L}$ is a bounded and strictly monotone operator;
• $\mathcal{L}$ is a mapping of $(S_{+})$ , i.e., if ${\phi }_n\rightharpoonup \phi$ inWand ${lim}_{n\to \mathrm{\infty }}sup\langle \mathcal{L}({\phi }_n)-\mathcal{L}(\phi ),{\phi }_n-\phi \rangle \le 0$ , then ${\phi }_n\to \phi$ inW;
• $\mathcal{L}:W\to W^{\ast }$ is a homeomorphism.

Define function $S:W\to \mathbb{R}$

$S(\phi )={\int }_{\mathrm{\Omega }\times \mathrm{\Omega }}\frac{|\phi (x)-\phi (y){|}^{p(x,y)}}{p(x,y)|x-y{|}^{N+sp(x,y)}}dxdy,\text{for all}\phi \in W,$

Graph

which is related to (9). The derivative of S is

$\langle S^{\prime }(\phi ),\psi \rangle ={\int }_{\mathrm{\Omega }}\frac{|\phi (x)-\phi (y){|}^{p(x,y)-2}(\phi (x)-\phi (y))(\psi (x)-\psi (y))}{|x-y{|}^{N+sp(x,y)}}dy=\langle \mathcal{L}(\phi ),\psi \rangle ,$

Graph

for all $\phi ,\psi \in W$ ; for more details, see [[34]].

Let

$X_i=\{\phi \in W:{\int }_{\mathrm{\Omega }}\frac{V_i(x)|\phi {|}^{{\bar{p}}_i(x)}}{{\mu }^{{\bar{p}}_i(x)}}dx<+\mathrm{\infty },\text{for some}\mu >0\},$

Graph

equipped with the norm

${\parallel \phi \parallel }_{X_i}:={[\phi ]}_{s,p_i(x,\cdot ),{\mathbb{R}}^{2N}\mathrm{\setminus }{({\mathrm{\Omega }}^c)}^2}+{[\phi ]}_{V_i(\mathrm{\Omega })}+{\parallel {\beta }^{\frac{1}{{\bar{p}}_i(x)}}\phi \parallel }_{L^{{\bar{p}}_i(x)}(C\mathrm{\Omega })}+{\parallel |g{|}^{\frac{1}{{\bar{p}}_i(x)}}\phi \parallel }_{L^{{\bar{p}}_i(x)}(C\mathrm{\Omega })},$

Graph

where

${[\phi ]}_{V_i(\mathrm{\Omega })}=inf\{\mu >0:{\int }_{\mathrm{\Omega }}{V}_i(x)\frac{|\phi (x){|}^{{\bar{p}}_i(x)}}{{\mu }^{{\bar{p}}_i(x)}}dx<1\},$

Graph

and

${[\phi ]}_{s,p_i(x,\cdot ),{\mathbb{R}}^{2N}\mathrm{\setminus }{({\mathrm{\Omega }}^c)}^2}=inf\{\mu \ge 0:\frac{1}{2}{\int }_{{\mathbb{R}}^{2N}\mathrm{\setminus }{({\mathrm{\Omega }}^c)}^2}\frac{|\phi (x)-\phi (y){|}^{p_i(x,y)}}{{\mu }^{p_i(x,y)}|x-y{|}^{N+s{p}_i(x,y)}}dxdy\le 1\}$

Graph

with ${\mathrm{\Omega }}^c={\mathbb{R}}^N\mathrm{\setminus }\mathrm{\Omega }$ .

Lemma 3.5

([[28]]) Assume that assumptions (P), (G), and (V) hold. Then, $(X_i,{\parallel \cdot \parallel }_{X_i})$ is a reflexive Banach space.

The norm ${\parallel \cdot \parallel }_{X_i}$ on $X_i$ is equivalent to

$\begin{array}{rl}{\parallel \phi \parallel }_{s,p_i(x,\cdot ),{\mathbb{R}}^{2N}\mathrm{\setminus }{({\mathrm{\Omega }}^c)}^2}=& inf\{\mu \ge 0|{\rho }_{s,p_i(x,\cdot ),{\mathbb{R}}^{2N}\mathrm{\setminus }{({\mathrm{\Omega }}^c)}^2}(\frac{\phi }{\mu }\le 1)\}\\ =& inf\{\mu \ge 0|{\int }_{{\mathbb{R}}^{2N}\mathrm{\setminus }{({\mathrm{\Omega }}^c)}^2}\frac{|\phi (x)-\phi (y){|}^{p_i(x,y)}}{{\mu }^{p_i(x,y)}{p}_i(x,y)|x-y{|}^{N+s{p}_i(x,y)}}dxdy\\ & +{\int }_{\mathrm{\Omega }}\frac{V_i(x)}{{\mu }^{{\bar{p}}_i(x)}{\bar{p}}_i(x)}|\phi {|}^{{\bar{p}}_i(x)}dx+{\int }_{C\mathrm{\Omega }}\frac{\beta (x)}{{\mu }^{{\bar{p}}_i(x)}{\bar{p}}_i(x)}|\phi {|}^{{\bar{p}}_i(x)}dx\\ & +{\int }_{C\mathrm{\Omega }}\frac{g(x)}{{\mu }^{{\bar{p}}_i(x)}{\bar{p}}_i(x)}|\phi {|}^{{\bar{p}}_{i}x)}dx\le 1\},\end{array}$

Graph

where the modular ${\rho }_{s,p_i(x,\cdot ),{\mathbb{R}}^{2N}\mathrm{\setminus }{({\mathrm{\Omega }}^c)}^2}:X_i\to \mathbb{R}$ is defined by

10 $\begin{array}{rl}{\rho }_{s,p_i(x,\cdot ),{\mathbb{R}}^{2N}\mathrm{\setminus }{({\mathrm{\Omega }}^c)}^2}=& {\int }_{{\mathbb{R}}^{2N}\mathrm{\setminus }{({\mathrm{\Omega }}^c)}^2}\frac{|\phi (x)-\phi (y){|}^{p_i(x,y)}}{p_i(x,y)|x-y{|}^{N+s{p}_i(x,y)}}dxdy+{\int }_{\mathrm{\Omega }}\frac{V_i(x)}{{\bar{p}}_i(x)}|\phi {|}^{{\bar{p}}_i(x)}dx\\ & +{\int }_{C\mathrm{\Omega }}\frac{\beta (x)}{{\bar{p}}_i(x)}|\phi {|}^{{\bar{p}}_i(x)}dx+{\int }_{C\mathrm{\Omega }}\frac{g(x)}{{\bar{p}}_i(x)}|\phi {|}^{{\bar{p}}_i(x)}dx.\end{array}$

Graph

Lemma 3.6

Assume that assumptions (P), (G), and (V) hold. The following properties hold:

• ${\parallel \phi \parallel }_{X_i}<1(=1,>1)⟺{\rho }_{s,p_i(x,\cdot ),{\mathbb{R}}^{2N}\mathrm{\setminus }{({\mathrm{\Omega }}^c)}^2}(\phi )<1(=1,>1)$ ;
• ${\parallel \phi \parallel }_{X_i}\ge 1⟹{\parallel \phi \parallel }_{X_i}^{p_i^{-}}\le {\rho }_{s,p_i(x,\cdot ),{\mathbb{R}}^{2N}\mathrm{\setminus }{({\mathrm{\Omega }}^c)}^2}(\phi )\le {\parallel \phi \parallel }_{X_i}^{p_i^{+}}$ ;
• ${\parallel \phi \parallel }_{X_i}\le 1⟹{\parallel \phi \parallel }_{X_i}^{p_i^{+}}\le {\rho }_{s,p_i(x,\cdot ),{\mathbb{R}}^{2N}\mathrm{\setminus }{({\mathrm{\Omega }}^c)}^2}(\phi )\le {\parallel \phi \parallel }_{X_i}^{p_i^{-}}$ ;
• ${\rho }_{s,p_i(x,\cdot ),{\mathbb{R}}^{2N}\mathrm{\setminus }{({\mathrm{\Omega }}^c)}^2}(\phi -v)\to 0\iff {\parallel \phi -v\parallel }_{X_i}\to 0$ .

Let $X=X_1\cap X_2$ with norm ${\parallel \phi \parallel }_X={\parallel \phi \parallel }_{X_1}+{\parallel \phi \parallel }_{X_2}$ , which is a separable and reflexive Banach space. The dual space of X is $X^{\ast }$ . The modular ${\rho }_{s,p(x,\cdot ),{\mathbb{R}}^{2N}\mathrm{\setminus }{({\mathrm{\Omega }}^c)}^2}={\rho }_{s,p_1(x,\cdot ),{\mathbb{R}}^{2N}\mathrm{\setminus }{({\mathrm{\Omega }}^c)}^2}+{\rho }_{s,p_2(x,\cdot ),{\mathbb{R}}^{2N}\mathrm{\setminus }{({\mathrm{\Omega }}^c)}^2}$ . We have the following result.

Lemma 3.7

([[28]]) Assume that assumptions (P), (G), and (V) hold. Then, from (10), the following properties hold:

• The function ${\rho }_{s,p(x,\cdot ),{\mathbb{R}}^{2N}\mathrm{\setminus }{({\mathrm{\Omega }}^c)}^2)}$ is of class $C^1(X,\mathbb{R})$ ;
• The strictly monotone operator ${\rho }_{s,p(x,\cdot ),{\mathbb{R}}^{2N}\mathrm{\setminus }{({\mathrm{\Omega }}^c)}^2}^{\prime }:X\to X^{\ast }$ is coercive, then
• $\frac{{\langle {\rho }_{s,p(x,\cdot ),{\mathbb{R}}^{2N}\mathrm{\setminus }{({\mathrm{\Omega }}^c)}^2}^{\prime },\phi \rangle }_X}{{\parallel \phi \parallel }_X}\to +\mathrm{\infty },{\parallel \phi \parallel }_X\to +\mathrm{\infty };$

Graph

• ${\rho }_{s,p(x,\cdot ),{\mathbb{R}}^{2N}\mathrm{\setminus }{({\mathrm{\Omega }}^c)}^2}^{\prime }$ is a mapping of type $(S_{+})$ , that is, if ${\phi }_n\rightharpoonup \phi$ inXand $lim{sup}_{n\to +\mathrm{\infty }}{\langle {\rho }_{s,p(x,\cdot ),{\mathbb{R}}^{2N}\mathrm{\setminus }{({\mathrm{\Omega }}^c)}^2}^{\prime },\phi \rangle }_X\le 0$ , then ${\phi }_n\to \phi$ inX.

Lemma 3.8

([[35]]) Assume that assumptions (P), (G), (V), and (H) hold. Then, for any $\hat{r}\in C_{+}(\bar{\mathrm{\Omega }})$ with $1<\hat{r}(x)<p_s^{\ast }(x)$ for all $x\in \mathrm{\Omega }$ , there is a positive constant ${\varpi }^{\ast }={\varpi }^{\ast }(s,p_i,N,\hat{r},\mathrm{\Omega })>0$ such that

${\parallel \phi \parallel }_{L^{\hat{r}(x)}(\mathrm{\Omega })}\le {\varpi }^{\ast }{\parallel \phi \parallel }_X,\mathit{\text{for all}}\phi \in X.$

Graph

Moreover, this embedding is compact.

Lemma 3.9

([[35]]) Assume that assumptions (P), (G), (V), and (H) hold. Then, for any $\hat{r}\in C_{+}({\mathbb{R}}^N\mathrm{\setminus }\mathrm{\Omega })$ with $1<\hat{r}(x)<p_{\partial }^{\ast }(x)$ for all $x\in {\mathbb{R}}^N\mathrm{\setminus }\mathrm{\Omega }$ , there is a positive constant ${\hat{\varpi }}^{\ast }={\hat{\varpi }}^{\ast }(s,p_i,N,\hat{r},\partial \mathrm{\Omega })>0$ such that

${\parallel \phi \parallel }_{L^{\hat{r}(x)}({\mathbb{R}}^N\mathrm{\setminus }\mathrm{\Omega })}\le {\hat{\varpi }}^{\ast }{\parallel \phi \parallel }_X,\mathit{\text{for all}}\phi \in X.$

Graph

Moreover, this embedding is compact.

More precisely, we now present the divergence theorem and the analogous formula for the partition integral formula in nonlocal case [[37]].

Lemma 3.10

([[29]]) Let the hypotheses (P) hold, and letφbe any bounded $C^2$ -function in $R^N$ . Then,

$\begin{array}{rl}& {\int }_{\mathrm{\Omega }}{(-\mathrm{\Delta })}_{p_1(x,\cdot )}^s\phi (x)dx+{\int }_{\mathrm{\Omega }}{(-\mathrm{\Delta })}_{p_2(x,\cdot )}^s\phi (x)dx\\ & =-({\int }_{{\mathbb{R}}^N\setminus \mathrm{\Omega }}{\mathcal{N}}_{s,p_1(x,\cdot )}\phi (x)dx+{\int }_{{\mathbb{R}}^N\setminus \mathrm{\Omega }}{\mathcal{N}}_{s,p_2(x,\cdot )}\phi (x)dx).\end{array}$

Graph

Lemma 3.11

Let the hypotheses (P) hold. Suppose thatφandvare bounded $C^2$ -functions in ${\mathbb{R}}^N$ . Then,

$\begin{array}{rl}& \frac{1}{2}\left({\int }_{{\mathbb{R}}^{2N}\setminus {(C\mathrm{\Omega })}^2}\right|\phi (x)-\phi (y){|}^{p_1(x,y)-2}\frac{(\phi (x)-\phi (y))(v(x)-v(y))}{|x-y{|}^{N+s{p}_1(x,y)}}dxdy\\ & +{\int }_{{\mathbb{R}}^{2N}\setminus {(C\mathrm{\Omega })}^2}|\phi (x)-\phi (y){|}^{p_2(x,y)-2}\frac{(\phi (x)-\phi (y))(v(x)-v(y))}{|x-y{|}^{N+s{p}_2(x,y)}}dxdy)\\ & ={\int }_{\mathrm{\Omega }}v{(-\mathrm{\Delta })}_{p_1(x,\cdot )}^s\phi (x)dx+{\int }_{\mathrm{\Omega }}v{(-\mathrm{\Delta })}_{p_2(x,\cdot )}^s\phi (x)dx\\ & +{\int }_{C\mathrm{\Omega }}v{\mathcal{N}}_{s,p_1(x,\cdot )}\phi (x)dx+{\int }_{C\mathrm{\Omega }}v{\mathcal{N}}_{s,p_2(x,\cdot )}\phi (x)dx,\end{array}$

Graph

for every $v\in X$ .

Proof

According to symmetry, we obtain

$\begin{array}{rl}& \frac{1}{2}\left({\int }_{{\mathbb{R}}^{2N}\setminus {(C\mathrm{\Omega })}^2}\right|\phi (x)-\phi (y){|}^{p_1(x,y)-2}\frac{(\phi (x)-\phi (y))(v(x)-v(y))}{|x-y{|}^{N+s{p}_1(x,y)}}dxdy\\ & +{\int }_{{\mathbb{R}}^{2N}\setminus {(C\mathrm{\Omega })}^2}|\phi (x)-\phi (y){|}^{p_2(x,y)-2}\frac{(\phi (x)-\phi (y))(v(x)-v(y))}{|x-y{|}^{N+s{p}_2(x,y)}}dxdy)\\ & ={\int }_{\mathrm{\Omega }}{\int }_{{\mathbb{R}}^N}v(x)|\phi (x)-\phi (y){|}^{p_1(x,y)-2}\frac{\phi (x)-\phi (y)}{|x-y{|}^{N+s{p}_1(x,y)}}dydx\\ & +{\int }_{\mathrm{\Omega }}{\int }_{{\mathbb{R}}^N}v(x)|\phi (x)-\phi (y){|}^{p_2(x,y)-2}\frac{\phi (x)-\phi (y)}{|x-y{|}^{N+s{p}_2(x,y)}}dydx\\ & +{\int }_{C\mathrm{\Omega }}{\int }_{\mathrm{\Omega }}v(x)|\phi (x)-\phi (y){|}^{p_1(x,y)-2}\frac{\phi (x)-\phi (y)}{|x-y{|}^{N+s{p}_1(x,y)}}dydx\\ & +{\int }_{C\mathrm{\Omega }}{\int }_{\mathrm{\Omega }}v(x)|\phi (x)-\phi (y){|}^{p_2(x,y)-2}\frac{\phi (x)-\phi (y)}{|x-y{|}^{N+s{p}_2(x,y)}}dydx.\end{array}$

Graph

□

Lemma 3.12

Assuming that assumption (P) holds and lettingφbe a weak solutions of equations (1), we have

$\begin{array}{rl}& {\mathcal{N}}_{s,p_1(x,\cdot )}\phi +{\mathcal{N}}_{s,p_2(x,\cdot )}\phi +\beta (x)|\phi {|}^{{\overline{p}}_1(x)-2}\phi +\beta (x)|\phi {|}^{{\overline{p}}_2(x)-2}\phi \\ & =g_1(x)+g_2(x),\mathit{\text{a.e. in}}{\mathbb{R}}^N\setminus \bar{\mathrm{\Omega }}.\end{array}$

Graph

Lemma 3.13

Assuming that assumptions (P), (G), (V), and (H) hold, let $I_{\lambda }:X\to \mathbb{R}$ be a energy function defined by

$\begin{array}{rl}{I}_{\lambda }(\phi )=& \frac{1}{2}{\int }_{{\mathbb{R}}^{2N}\setminus {(C\mathrm{\Omega })}^2}\frac{|\phi (x)-\phi (y){|}^{p_1(x,y)}}{p_1(x,y)|x-y{|}^{N+s{p}_1(x,y)}}dxdy\\ & +\frac{1}{2}{\int }_{{\mathbb{R}}^{2N}\setminus {(C\mathrm{\Omega })}^2}\frac{|\phi (x)-\phi (y){|}^{p_2(x,y)}}{p_2(x,y)|x-y{|}^{N+s{p}_2(x,y)}}dxdy\\ & +{\int }_{\mathrm{\Omega }}{V}_1(x)\frac{|\phi {|}^{{\overline{p}}_1(x)}}{{\overline{p}}_1(x)}dx+{\int }_{\mathrm{\Omega }}{V}_2(x)\frac{|\phi {|}^{{\overline{p}}_2(x)}}{{\overline{p}}_2(x)}dx+{\int }_{{\mathbb{R}}^N\setminus \mathrm{\Omega }}\beta (x)\frac{|\phi {|}^{{\overline{p}}_1(x)}}{{\overline{p}}_1(x)}dx\\ & +{\int }_{{\mathbb{R}}^N\setminus \mathrm{\Omega }}\beta (x)\frac{|\phi {|}^{{\overline{p}}_2(x)}}{{\overline{p}}_2(x)}dx-{\lambda }_1{\int }_{\mathrm{\Omega }}{A}_1(x)\frac{|\phi {|}^{r_1(x)}}{r_1(x)}dx-{\lambda }_2{\int }_{\mathrm{\Omega }}{A}_2(x)\frac{|\phi {|}^{r_2(x)}}{r_2(x)}dx\\ & -{\int }_{{\mathbb{R}}^N\setminus \mathrm{\Omega }}{g}_1(x)\phi dx-{\int }_{{\mathbb{R}}^N\setminus \mathrm{\Omega }}{g}_2(x)\phi dx,\end{array}$

Graph

for every $\phi \in X$ . Then, any critical point of $I_{\lambda }$ is a weak solution of equations (1).

Proof of Theorem 1.1

To prove Theorem 1.1, we need a well-known mountain pass lemma.

Theorem 4.1

Let X be a real Banach space and $I_{\lambda }\in C^1(X,\mathbb{R})$ with $I_{\lambda }(0)=0$ . Assume that the following conditions hold:

• $I_{\lambda }$ satisfies $(PS)$ conditions;
• there exist $\rho ,\sigma >0$ such that $I_{\lambda }(\phi )\ge \sigma$ , for all $\phi \in X$ , with ${\parallel \phi \parallel }_X=\rho$ ;
• there exists $\upsilon \in X$ , satisfying ${\parallel \upsilon \parallel }_X>\rho$ such that $I_{\lambda }(\upsilon )<0$ .

Then, $I_{\lambda }$ has a critical value $c>\sigma$ , that is,

$c=\underset{\gamma \in \mathrm{\Upsilon }}{inf}\underset{0\le t\le 1}{max}{I}_{\lambda }(\gamma (t)),$

Graph

where $\mathrm{\Upsilon }=\{\gamma \in C^1([0,1];X):\gamma (0)=1,\gamma (1)=\upsilon \}$ .

Definition 1

Let X be a Banach space, $I_{\lambda }\in C^1(E,\mathbb{R})$ . We say that $I_{\lambda }$ satisfies the $(PS)$ conditions if every sequence ${\{{\phi }_n\}}_{n\in \mathbb{N}}\subset X$ satisfying

$I_{\lambda }({\phi }_n)\to c,I_{\lambda }({\phi }_n)\to 0,n\to \mathrm{\infty }$

Graph

has a convergent subsequence in X.

Next, we prove that the $I_{\lambda }$ defined in Lemma 3.13 satisfies the $(PS)$ conditions.

Lemma 4.1

Assume that assumptions (P), (G), (V), and (H) hold. Then, the sequence ${\{{\phi }_n\}}_{n\in \mathbb{N}}$ is bounded inX.

Proof

According to (H), we get

$s_1^{\prime }(x)r_1(x)<p_s^{\ast }(x),s_2^{\prime }(x)r_2(x)<p_s^{\ast }(x),\text{for all}x\in \mathrm{\Omega },$

Graph

so from Lemmas 3.8 and 3.9, there exist constants $M_1$ and $M_2$ such that

11 ${\parallel \phi \parallel }_{L^{s_1^{\prime }(x)r_1(x)}(\mathrm{\Omega })}\le M_1{\parallel \phi \parallel }_X,{\parallel \phi \parallel }_{L^{s_2^{\prime }(x)r_2(x)}(\mathrm{\Omega })}\le M_2{\parallel \phi \parallel }_X,\text{for all}\phi \in X.$

Graph

Let $\rho >max\{1,\frac{1}{M_1},\frac{1}{M_2}\}$ and

${\parallel \phi \parallel }_{L^{s_1^{\prime }(x)r_1(x)}(\mathrm{\Omega })}>1,{\parallel \phi \parallel }_{L^{s_2^{\prime }(x)r_2(x)}(\mathrm{\Omega })}>1.$

Graph

Thus, by the Hölder inequality and Lemma 2.3, for all $\phi \in X$ with ${\parallel \phi \parallel }_X=\rho$ , we obtain

12 $\begin{array}{rl}{\int }_{\mathrm{\Omega }}{A}_1(x)|\phi (x){|}^{r_1(x)}dx& \le 2{\parallel A_1\parallel }_{s_1(x)}{\parallel \phi \parallel }_{s_1^{\prime }(x)r_1(x)}^{r_1^{+}}\le 2M_1^{r_1^{+}}{\parallel A_1\parallel }_{s_1(x)}{\parallel \phi \parallel }_X^{r_1^{+}}\end{array}$

Graph

and

13 $\begin{array}{rl}{\int }_{\mathrm{\Omega }}{A}_2(x)|\phi (x){|}^{r_2(x)}dx& \le 2{\parallel A_2\parallel }_{s_2(x)}{\parallel \phi \parallel }_{s_2^{\prime }(x)r_2(x)}^{r_2^{+}}\le 2M_2^{r_2^{+}}{\parallel A_2\parallel }_{s_2(x)}{\parallel \phi \parallel }_X^{r_2^{+}}.\end{array}$

Graph

We use the counterfactual method. Suppose ${\parallel {\phi }_n\parallel }_X\to \mathrm{\infty }$ , $n\to \mathrm{\infty }$ . Combining conditions (P), (G), (V), (H), and Lemma 3.8 and letting $0<\vartheta <min\{r_1^{-}{r}_2^{-}\frac{Z_A+Z_B-1}{r_2^{-}{Z}_A{K}_A+r_1^{-}{Z}_B{K}_A},r_1^{-},r_2^{-}\}$ , where $K_A=max\{{\parallel A_1\parallel }_X,{\parallel A_2\parallel }_X\}$ , $Z_A=4{\lambda }_1{M}_1^{r_1^{+}}{C}_{A_1}$ , $Z_B=4{\lambda }_2{M}_2^{r_2^{+}}{C}_{A_2}$ , we have

$\begin{array}{rl}& c\frac{1}{{\parallel {\phi }_n\parallel }_X^{p_1^{-}}}+o(1)\frac{1}{{\parallel {\phi }_n\parallel }_X^{p_1^{-}}}\\ & \ge \frac{1}{{\parallel {\phi }_n\parallel }_X^{p_1^{-}}}(I_{\lambda }({\phi }_n)-\langle \frac{1}{\vartheta }{I}_{\lambda }^{\prime }({\phi }_n),{\phi }_n\rangle )\\ & =\frac{1}{{\parallel {\phi }_n\parallel }_X^{p_1^{-}}}\sum _{i=1}^2[{\int }_{{\mathbb{R}}^{2N}\setminus {(\mathcal{C}\mathrm{\Omega })}^2}\frac{|\phi (x)-\phi (y){|}^{p_i(x,y)}}{2p_i(x,y)|x-y{|}^{N+s{p}_i(x,y)}}dxdy+{\int }_{\mathrm{\Omega }}{V}_i(x)\frac{|{\phi }_n{|}^{{\overline{p}}_i(x)}}{{\overline{p}}_i(x)}dx\\ & +{\int }_{{\mathbb{R}}^N\setminus \mathrm{\Omega }}\beta (x)\frac{|{\phi }_n{|}^{{\overline{p}}_i(x)}}{{\overline{p}}_i(x)}dx-{\lambda }_1{\int }_{\mathrm{\Omega }}{A}_1(x)\frac{|{\phi }_n{|}^{r_1(x)}}{r_1(x)}dx-{\lambda }_2{\int }_{\mathrm{\Omega }}{A}_2(x)\frac{|{\phi }_n{|}^{r_2(x)}}{r_2(x)}dx\\ & -{\int }_{{\mathbb{R}}^N\setminus \mathrm{\Omega }}{g}_i(x){\phi }_{n}dx-\frac{1}{\vartheta }({\int }_{{\mathbb{R}}^{2N}\setminus {(\mathcal{C}\mathrm{\Omega })}^2}\frac{|{\phi }_n(x)-{\phi }_n(y){|}^{p_i(x,y)}}{2|x-y{|}^{N+s{p}_i(x,y)}}dxdy\\ & +{\int }_{\mathrm{\Omega }}{V}_i(x)|{\phi }_n{|}^{{\overline{p}}_i(x)}dx+{\int }_{{\mathbb{R}}^N\setminus \mathrm{\Omega }}\beta (x)|{\phi }_n{|}^{{\overline{p}}_i(x)}dx-{\lambda }_1{\int }_{\mathrm{\Omega }}{A}_1(x)|{\phi }_n{|}^{r_1(x)}dx\\ & -{\lambda }_2{\int }_{\mathrm{\Omega }}{A}_2(x)|{\phi }_n{|}^{r_2(x)}dx-{\int }_{{\mathbb{R}}^N\setminus \mathrm{\Omega }}{g}_i(x){\phi }_{n}dx)]\\ & \ge \frac{1}{{\parallel {\phi }_n\parallel }_X^{p_1^{-}}}\left[(\frac{1}{2p_2^{+}}-\frac{1}{2\vartheta })\sum _{i=1}^2\right({\int }_{{\mathbb{R}}^{2N}\setminus {(\mathcal{C}\mathrm{\Omega })}^2}\frac{|{\phi }_n(x)-{\phi }_n(y){|}^{p_i(x,y)}}{|x-y{|}^{N+s{p}_i(x,y)}}dxdy\\ & +{\int }_{\mathrm{\Omega }}{V}_i(x)|{\phi }_n{|}^{{\overline{p}}_i(x)}dx+{\int }_{{\mathbb{R}}^N\setminus \mathrm{\Omega }}\beta (x)|{\phi }_n{|}^{{\overline{p}}_i(x)}dx-{\int }_{{\mathbb{R}}^N\setminus \mathrm{\Omega }}{g}_i(x){\phi }_{n}dx)\\ & -2(\vartheta -r_1^{+})\frac{{\lambda }_1{M}_1^{r_1^{+}}}{\vartheta r_1^{-}}{\parallel A_1\parallel }_{s_1(x)}{\parallel {\phi }_n\parallel }_X^{r_1^{+}}-2(\vartheta -r_2^{+})\frac{{\lambda }_2{M}_2^{r_2^{+}}}{\vartheta r_2^{-}}{\parallel A_2\parallel }_{s_2(x)}{\parallel {\phi }_n\parallel }_X^{r_2^{+}}]\\ & \ge (\frac{1}{2p_2^{+}}-\frac{1}{2\vartheta })-2(\vartheta -r_1^{-})\frac{{\lambda }_1{M}_1^{r_1^{+}}{C}_{A_1}}{\vartheta r_1^{-}}{\parallel A_1\parallel }_X-2(\vartheta -r_2^{-})\frac{{\lambda }_2{M}_2^{r_2^{+}}{C}_{A_2}}{\vartheta r_2^{-}}{\parallel A_2\parallel }_X.\end{array}$

Graph

In addition, we obtain $c\frac{1}{{\parallel {\phi }_n\parallel }_X^{p_1^{-}}}+o(1)\frac{1}{{\parallel {\phi }_n\parallel }_X^{p_1^{-}}}\to 0$ because ${\parallel {\phi }_n\parallel }_X\to \mathrm{\infty }$ , $n\to \mathrm{\infty }$ . Due to

$0<\vartheta <min\{r_1^{-}{r}_2^{-}\frac{Z_A+Z_B-1}{r_2^{-}{Z}_A{K}_A+r_1^{-}{Z}_B{K}_A},r_1^{-},r_2^{-}\},$

Graph

there is a contradiction. Thus, ${\{{\phi }_n\}}_{n\in \mathbb{N}}$ is bounded. □

Inspired by [[15]], we have the following lemma.

Lemma 4.2

Assume that assumptions (P), (G), (V), and (H) hold. Then, $I_{\lambda }$ satisfies the $(PS)$ conditions.

Proof

According to Lemma 4.1, ${\{{\phi }_n\}}_{n\in \mathbb{N}}$ is bounded, that is, there is a subsequence ${\{{\phi }_n\}}_{n\in \mathbb{N}}$ and ${\phi }_0$ in X such that

$\begin{array}{rl}& {\phi }_n\rightharpoonup {\phi }_0\text{in}X;\\ & {\phi }_n\to {\phi }_0\text{a.e. in}\mathrm{\Omega };\\ & {\phi }_n\to {\phi }_0\text{in}{L}^{\hat{r}(x)}(\mathrm{\Omega }),\hat{r}(x)<p_s^{\ast }(x);\\ & {\phi }_n\to {\phi }_0\text{in}{L}^{\hat{r}(x)}({\mathbb{R}}^N\setminus \mathrm{\Omega }),\hat{r}(x)<p_{\partial }^{\ast }(x).\end{array}$

Graph

Due to ${\phi }_n\to {\phi }_0$ in $L^{\hat{r}(x)}({\mathbb{R}}^N\setminus \mathrm{\Omega })$ , then ${|{\phi }_n|}^{{\overline{p}}_i(x)-2}{\phi }_n\to {|{\phi }_0|}^{{\overline{p}}_i(x)-2}{\phi }_0$ . We get

$\underset{n\to \mathrm{\infty }}{lim}{\int }_{\mathbb{R}\setminus \mathrm{\Omega }}\beta (x)(|{\phi }_n{|}^{{\overline{p}}_i(x)-2}{\phi }_n-|{\phi }_0{|}^{{\overline{p}}_i(x)-2}{\phi }_0)=0.$

Graph

Since ${\lambda }_1{A}_1(x){|\phi |}^{r_1(x)-2}\phi$ and ${\lambda }_2{A}_2(x){|\phi |}^{r_2(x)-2}\phi$ in X are sequentially weakly lower semi-continuous, for $v\in X$ and measurable for all $\mathrm{\Omega }\subset {\mathbb{R}}^N$ , we obtain

$\begin{array}{rl}& |{\int }_{\mathrm{\Omega }}{A}_1(x)(|\phi {|}_n^{r_1(x)-2}{\phi }_n-|\phi {|}_0^{r_1(x)-2}{\phi }_0)|v|dx|\\ & \le {\int }_{\mathrm{\Omega }}{A}_1(x)(|\phi {|}_n^{r_1(x)-1}-|\phi {|}_0^{r_1(x)-1})vdx\\ & ={\int }_{\mathrm{\Omega }}{A}_1{(x)}^{\frac{r_1(x)-1}{r_1(x)}}(|\phi {|}_n^{r_1(x)-1}-|\phi {|}_0^{r_1(x)-1})A_1{(x)}^{\frac{1}{r_1(x)}}|v|dx\\ & \le {\parallel A_1{(x)}^{\frac{r_1(x)-1}{r_1(x)}}(|\phi {|}_n^{r_1(x)-1}-|\phi {|}_0^{r_1(x)-1}\parallel }_{L^{\frac{r_1(x)}{r_1(x)-1}}}{\parallel A_1{(x)}^{\frac{1}{r_1(x)}})|v|\parallel }_{L^{r_1(x)}}.\end{array}$

Graph

Hence, ${\{A_1(x)({|\phi |}_n^{r_1(x)-2}{\phi }_n-{|\phi |}_0^{r_1(x)-2}{\phi }_0)|v|\}}_{n\in \mathbb{N}}$ is uniformly integrable in ${\mathbb{R}}^N$ . Then, using the Vitali convergence theorem, we get

$\underset{n\to \mathrm{\infty }}{lim}{\int }_{\mathrm{\Omega }}{A}_1(x)(|{\phi }_n{|}^{r_1(x)-2}{\phi }_n-|{\phi }_0{|}^{r_1(x)-2}{\phi }_0)=0.$

Graph

Similarly, there is

$\underset{n\to \mathrm{\infty }}{lim}{\int }_{\mathrm{\Omega }}{A}_2(x)(|{\phi }_n{|}^{r_2(x)-2}{\phi }_n-|{\phi }_0{|}^{r_2(x)-2}{\phi }_0)=0.$

Graph

We need to prove that ${\{{\phi }_n\}}_{n\in \mathbb{N}}$ is strongly convergent,

$\begin{array}{rl}o(1)=& \langle I_{\lambda }^{\prime }({\phi }_n)-I_{\lambda }^{\prime }({\phi }_0),{\phi }_n-{\phi }_0\rangle \\ =& \sum _{i=1}^2[{\int }_{{\mathbb{R}}^{2N}\setminus {(\mathcal{C}\mathrm{\Omega })}^2}\frac{|{\phi }_n(x)-{\phi }_n(y){|}^{p_i(x,y)-2}({\phi }_n(x)-{\phi }_n(y))({\phi }_n(x)-{\phi }_n(y)-{\phi }_0(x)+{\phi }_0(y))}{2|x-y{|}^{N+s{p}_i(x,y)}}dxdy\\ & -{\int }_{{\mathbb{R}}^{2N}\setminus {(\mathcal{C}\mathrm{\Omega })}^2}\frac{|{\phi }_0(x)-{\phi }_0(y){|}^{p_i(x,y)-2}({\phi }_0(x)-{\phi }_0(y))({\phi }_n(x)-{\phi }_n(y)-{\phi }_0(x)+{\phi }_0(y))}{2|x-y{|}^{N+s{p}_i(x,y)}}dxdy\\ & +{\int }_{\mathrm{\Omega }}{V}_i(x)(|{\phi }_n{|}^{{\overline{p}}_i(x)-2}{\phi }_n-|{\phi }_0{|}^{{\overline{p}}_i(x)-2}{\phi }_0)({\phi }_n-{\phi }_0)dx\\ & +{\int }_{{\mathbb{R}}^N\setminus \mathrm{\Omega }}\beta (x)(|{\phi }_n{|}^{{\overline{p}}_i(x)-2}{\phi }_n-|{\phi }_0{|}^{{\overline{p}}_i(x)-2}{\phi }_0)({\phi }_n-{\phi }_0)dx\\ & -{\lambda }_1{\int }_{\mathrm{\Omega }}{A}_1(x)(|{\phi }_n{|}^{{\overline{p}}_i(x)-2}{\phi }_n-|{\phi }_0{|}^{{\overline{p}}_i(x)-2}{\phi }_0)({\phi }_n-{\phi }_0)dx\\ & -{\lambda }_2{\int }_{\mathrm{\Omega }}{A}_2(x)(|{\phi }_n{|}^{{\overline{p}}_i(x)-2}{\phi }_n-|{\phi }_0{|}^{{\overline{p}}_i(x)-2}{\phi }_0)({\phi }_n-{\phi }_0)dx]\\ =& \sum _{i=1}^2[{\int }_{{\mathbb{R}}^{2N}\setminus {(\mathcal{C}\mathrm{\Omega })}^2}\frac{|{\phi }_n(x)-{\phi }_n(y){|}^{p_i(x,y)-2}({\phi }_n(x)-{\phi }_n(y))({\phi }_n(x)-{\phi }_n(y)-{\phi }_0(x)+{\phi }_0(y))}{2|x-y{|}^{N+s{p}_i(x,y)}}dxdy\\ & -{\int }_{{\mathbb{R}}^{2N}\setminus {(\mathcal{C}\mathrm{\Omega })}^2}\frac{|{\phi }_0(x)-{\phi }_0(y){|}^{p_i(x,y)-2}({\phi }_0(x)-{\phi }_0(y))({\phi }_n(x)-{\phi }_n(y)-{\phi }_0(x)+{\phi }_0(y))}{2|x-y{|}^{N+s{p}_i(x,y)}}dxdy].\end{array}$

Graph

A discussion similar to Lemma 3.7 gives that ${\phi }_n\to {\phi }_0$ in X. Combining the Definition 1 and the Lemma 4.1, we complete the proof. □

Lemma 4.3

Assume that assumptions (P), (G), (V), and (H) hold. There exist $\rho >0$ and $\sigma >0$ such that, for all $\phi \in X$ with ${\parallel \phi \parallel }_X=\rho$ ,

$I_{\lambda }(\phi )\ge \sigma >0$

Graph

holds.

Proof

Combining (12) with (13), for any $\phi \in X$ with ${\parallel \phi \parallel }_X=\rho >1$ , we have

$\begin{array}{rl}{I}_{\lambda }(\phi )\ge & \frac{1}{p_2^{+}}\sum _{i=1}^2({\int }_{{\mathbb{R}}^{2N}\setminus {(\mathcal{C}\mathrm{\Omega })}^2}\frac{|\phi (x)-\phi (y){|}^{p_i(x,y)}}{2|x-y{|}^{N+s{p}_i(x,y)}}dxdy+{\int }_{\mathrm{\Omega }}{V}_i(x)|\phi {|}^{{\overline{p}}_i(x)}dx\\ & +{\int }_{{\mathbb{R}}^N\setminus \mathrm{\Omega }}\beta (x)|\phi {|}^{{\overline{p}}_i(x)}dx-{\int }_{{\mathbb{R}}^N\setminus \mathrm{\Omega }}{g}_i(x)\phi dx)\\ & -\frac{{\lambda }_1}{r_1^{-}}{\int }_{\mathrm{\Omega }}{A}_1(x)|\phi {|}^{r_1(x)}dx-\frac{{\lambda }_2}{r_2^{-}}{\int }_{\mathrm{\Omega }}{A}_2(x)|\phi {|}^{r_2(x)}dx\\ \ge & \frac{1}{p_2^{+}}{\parallel \phi \parallel }_X^{p_2^{+}}-{\lambda }_1\frac{2M_1^{r_1^{-}}{\parallel A_1\parallel }_{s_1(x)}}{r_1^{-}}{\parallel \phi \parallel }_X^{r_1^{-}}-{\lambda }_2\frac{2M_2^{r_2^{-}}{\parallel A_2\parallel }_{s_2(x)}}{r_2^{-}}{\parallel \phi \parallel }_X^{r_2^{-}}\\ =& {\parallel \phi \parallel }_X^{r_1^{-}}(\frac{1}{p_2^{+}}{\parallel \phi \parallel }_X^{p_2^{+}-r_1^{-}}-{\lambda }_2\frac{2M_2^{r_2^{-}}{\parallel A_2\parallel }_{s_2(x)}}{r_2^{-}}{\parallel \phi \parallel }_X^{r_2^{-}-r_1^{-}}-{\lambda }_1\frac{2M_1^{r_1^{-}}{\parallel A_1\parallel }_{s_1(x)}}{r_1^{-}}).\end{array}$

Graph

Let

$f(t)=\frac{1}{p_2^{+}}{t}^{p_2^{+}-r_1^{-}}-{\lambda }_2\frac{2M_2^{r_2^{-}}{\parallel A_2\parallel }_{s_2(x)}}{r_2^{-}}{t}^{r_2^{-}-r_1^{-}},t\ge 0,$

Graph

where $r_1^{-}<r_2^{-}<p_2^{+}$ . Then, there exists $\chi >0$ such that $f(t)=f(\chi )>0$ . Choosing ${\parallel A_1\parallel }_{s_1(x)}^{r_1^{-}}<{\sigma }^{\ast }=\frac{r_1^{-}}{4{\lambda }_1{M}_1^{r_1^{-}}}f(\chi )$ , we get

$I_{\lambda }(\phi )\ge \sigma ={\chi }^{r_1^{-}}\frac{f(\chi )}{2}>0,$

Graph

for ${\parallel \phi \parallel }_X=\chi =\rho$ . □

Lemma 4.4

Assume that assumptions (P), (G), (V), and (H) hold. Then, there existsυ, which satisfies ${\parallel \upsilon \parallel }_X>\rho$ . Then, there exists $\upsilon \in X$ such that

$I_{\lambda }(\upsilon )<0.$

Graph

Proof

Choosing $\hat{\upsilon }\in X$ such that ${\parallel \hat{\upsilon }\parallel }_X=1$ , and for $t\in (0,1)$ small enough, we obtain

$\begin{array}{rl}{I}_{\lambda }(t\hat{\upsilon })\le & t^{p_2^{+}}\sum _{i=1}^2({\int }_{{\mathbb{R}}^{2N}\setminus {(\mathcal{C}\mathrm{\Omega })}^2}\frac{|\hat{\upsilon }(x)-\hat{\upsilon }(y){|}^{p_i(x,y)}}{2p_i(x,y)|x-y{|}^{N+s{p}_i(x,y)}}dxdy+{\int }_{\mathrm{\Omega }}{V}_i(x)\frac{|\hat{\upsilon }{|}^{{\overline{p}}_i(x)}}{{\overline{p}}_i(x)}dx\\ & +{\int }_{{\mathbb{R}}^N\setminus \mathrm{\Omega }}\frac{\beta (x)|\hat{\upsilon }{|}^{{\overline{p}}_i(x)}}{{\overline{p}}_i(x)}dx)-{\lambda }_1{t}^{{r_1}^{-}}{\int }_{\mathrm{\Omega }}\frac{A_1(x)|\hat{\upsilon }{|}^{r_1(x)}}{r_1(x)}dx\\ & -{\lambda }_2{t}^{{r_2}^{-}}{\int }_{\mathrm{\Omega }}\frac{A_2(x)|\hat{\upsilon }{|}^{r_2(x)}}{r_2(x)}dx-{\int }_{{\mathbb{R}}^N\setminus \mathrm{\Omega }}t\hat{\upsilon }{g}_1(x)dx-{\int }_{{\mathbb{R}}^N\setminus \mathrm{\Omega }}t\hat{\upsilon }{g}_2(x)dx<0\end{array}$

Graph

with the fact that $1<r_1^{-}<p_2^{+}$ . Thus, $I_{\lambda }(t\hat{\upsilon })<0$ with ${\parallel t\hat{\upsilon }\parallel }_X>\rho$ . The proof is proved by letting $\upsilon =t\hat{\upsilon }$ . □

Proof of Theorem 1.1

Combining Lemmas 4.1 and 4.2, it can be inferred that $I_{\lambda }$ satisfies $(PS)$ conditions. According to Lemmas 4.3 and 4.4, we know that $I_{\lambda }$ satisfies the mountain pass lemma. Therefore, we have a subsequence ${\{{\phi }_n\}}_{n\in \mathbb{N}}$ and ${\phi }_0^{(1)}\in X$ such that ${\phi }_n\to {\phi }_0^{(1)}$ in X by Lemma 4.1 and $0<\sigma <c<\mathrm{\infty }$ . Therefore, $I_{\lambda }({\phi }_n)=c>\sigma$ , that is, ${\phi }_0^{(1)}$ is a solution of problem (1) with positive energy.

Next, we will apply Ekeland's variational principle to prove that (1) has a solution with negative energy.

By Lemma 4.3, we derive that

$\hat{c}=\underset{\partial B_{\rho }(0)}{inf}{I}_{\lambda }>0,$

Graph

where ρ is the positive constant introduced in Lemma 4.3.

From condition (H), there exist ${ϵ}_1,{ϵ}_2>0$ and an open set ${\mathrm{\Omega }}_0\subset \subset \mathrm{\Omega }$ such that

$|r_1(x)-{r_1}^{-}|<{ϵ}_1,|r_2(x)-{r_2}^{-}|<{ϵ}_2,\text{for all}x\in {\mathrm{\Omega }}_0,$

Graph

and we get

${r_1}^{-}+{ϵ}_1<p_1^{-},{r_2}^{-}+{ϵ}_2<p_1^{-},\text{for all}x\in {\bar{\mathrm{\Omega }}}_0.$

Graph

Hence,

14 $r_1(x)\le {r_1}^{-}+{ϵ}_1<p_1^{-},r_2(x)\le {r_2}^{-}+{ϵ}_2<p_1^{-},\text{for all}x\in {\bar{\mathrm{\Omega }}}_0.$

Graph

By Lemma 2.1 and $g_i(x)>0$ , we conclude

$\begin{array}{rl}{\int }_{\mathbb{R}\setminus \mathrm{\Omega }}{g}_i\phi dx\le & {\int }_{\mathbb{R}\setminus \mathrm{\Omega }}|g_i{|}^{\frac{1}{{\bar{p}}_i^{\prime }(x)}}|g_i{|}^{\frac{1}{{\bar{p}}_i(x)}}|\phi |dx\\ \le & 2{\parallel g_i\parallel }_{L^1(\mathbb{R}\setminus \mathrm{\Omega })}{\parallel |g_i{|}^{\frac{1}{{\bar{p}}_i(x)}}|\phi |\parallel }_{L^{{\bar{p}}_i(x)}(\mathbb{R}\setminus \mathrm{\Omega })}\le K_1{\parallel \phi \parallel }_X.\end{array}$

Graph

For sufficiently small $\tau \in (0,1)$ , let $\eta \in C_0^{\mathrm{\infty }}(\mathrm{\Omega })$ such that ${\bar{\mathrm{\Omega }}}_0\subset supp(\eta )$ , $\eta =1$ , for all $x\in {\bar{\mathrm{\Omega }}}_0$ and $0\le \eta \le 1$ in Ω. Then, by applying (14), it follows that

$\begin{array}{rl}{I}_{\lambda }(\tau \eta )\le & {\tau }^{p_1^{-}}\sum _{i=1}^2({\int }_{{\mathbb{R}}^{2N}\setminus {(\mathcal{C}\mathrm{\Omega })}^2}\frac{|\eta (x)-\eta (y){|}^{p_i(x,y)}}{2p_i(x,y)|x-y{|}^{N+s{p}_i(x,y)}}dxdy+{\int }_{\mathrm{\Omega }}\frac{V_i(x)|\eta {|}^{{\overline{p}}_i(x)}}{{\overline{p}}_i(x)}dx\\ & +{\int }_{{\mathbb{R}}^N\setminus \mathrm{\Omega }}\frac{\beta (x)|\eta {|}^{{\overline{p}}_i(x)}}{{\overline{p}}_i(x)}dx)-{\lambda }_1{\tau }^{{r_1}^{-}+{ϵ}_1}{\int }_{{\mathrm{\Omega }}_0}\frac{A_1(x)|\eta {|}^{r_1(x)}}{r_1(x)}dx\\ & -{\lambda }_2{\tau }^{{r_2}^{-}+{ϵ}_2}{\int }_{{\mathrm{\Omega }}_0}\frac{A_2(x)|\eta {|}^{r_2(x)}}{r_2(x)}dx-{\int }_{{\mathbb{R}}^N\setminus \mathrm{\Omega }}\tau \eta g_1(x)dx-{\int }_{{\mathbb{R}}^N\setminus \mathrm{\Omega }}\tau \eta g_2(x)dx\\ \le & \frac{{\tau }^{p_1^{-}}}{p_1^{-}}3K_{1}max\{{\parallel \eta \parallel }_X^{p_2^{-}},{\parallel \eta \parallel }_X^{p_2^{+}}\}-\frac{{\lambda }_2{\tau }^{{r_2}^{-}+{ϵ}_2}}{r_2^{+}}{\int }_{{\mathrm{\Omega }}_0}{A}_2(x)|\eta {|}_X^{r_2(x)}dx.\end{array}$

Graph

Since ${r_2}^{-}+{ϵ}_2<p_1^{-}$ , we have $I_{\lambda }(t\eta )<0$ .

In addition, combining the Hölder inequality and inequality (11), for any $\phi \in B_{\rho }(0)$ , we have

$I_{\lambda }(\phi )\ge \frac{1}{p_2^{+}}{\parallel \phi \parallel }_X^{p_1^{-}}-\frac{2{\lambda }_1{M}_1^{r_1^{-}}}{r_1^{-}}{\parallel A_1\parallel }_{s_1(x)}{\parallel \phi \parallel }_X^{r_1^{-}}-\frac{2{\lambda }_2{M}_2^{r_2^{-}}}{r_2^{-}}{\parallel A_2\parallel }_{s_2(x)}{\parallel \phi \parallel }_X^{r_2^{-}}-2K_1{\parallel \phi \parallel }_X.$

Graph

This fact gives

15 $-\mathrm{\infty }<\hat{c}:={\underset{\_}{inf}}_{B_{\rho }(0)}{I}_{\lambda }<0.$

Graph

Set

$\frac{1}{n}\in (0,{\underset{\_}{inf}}_{\partial B_{\rho }(0)}{I}_{\lambda }(\phi )-{\underset{\_}{inf}}_{B_{\rho }(0)}{I}_{\lambda }(\phi )),n\in \mathbb{N}.$

Graph

By (15), $I_{\lambda }:\overline{B_{\rho }(0)}\to \mathbb{R}$ is lower bounded on $\overline{B_{\rho }(0)}$ and $I_{\lambda }\in C^1(\overline{B_{\rho }(0)},\mathbb{R})$ . Using Ekeland's variational principle, there exists ${\{{\phi }_n\}}_{n\in \mathbb{N}}\in \overline{B_{\rho }(0)}$ such that

16 $\{\begin{array}{c}\hat{c}\le I_{\lambda }({\phi }_n)\le \hat{c}+\frac{1}{n},\hfill \\ 0<I_{\lambda }({\phi }_n)-I_{\lambda }(\phi )+\frac{1}{n}{\parallel {\phi }_n-\phi \parallel }_X,{\phi }_n\ne \phi .\hfill \end{array}$

Graph

Since

$I_{\lambda }({\phi }_n)\le \hat{c}+\frac{1}{n}\le \underset{\overline{B_{\rho }(0)}}{inf}{I}_{\lambda }+\frac{1}{n}\le \underset{B_{\rho }(0)}{inf}{I}_{\lambda }+\frac{1}{n}<\underset{\partial B_{\rho }(0)}{inf}{I}_{\lambda },$

Graph

we have ${\phi }_n\in B_{\rho }(0)$ . Define function $\zeta :\overline{B_{\rho }(0)}\to \mathbb{R}$ by

$\zeta (\phi )=I_{\lambda }(\phi )+\frac{1}{n}\cdot {\parallel {\phi }_n-\phi \parallel }_X,$

Graph

which implies $\zeta ({\phi }_n)<\zeta (\phi )$ from (16). Then, ${\phi }_n$ is a minimum point of ζ, and we have

$\frac{\zeta ({\phi }_n+t\cdot v)-\zeta ({\phi }_n)}{t}\ge 0,$

Graph

for small $t>0$ and any $v\in B_1(0)=\{v\in X:{\parallel v\parallel }_X=1\}$ . Hence,

$\frac{I_{\lambda }({\phi }_n+t\cdot v)-I_{\lambda }({\phi }_n)}{t}+\frac{1}{n}{\parallel v\parallel }_X\ge 0.$

Graph

Let $t\to 0$ , then $\langle I_{\lambda }^{\prime },v\rangle +\frac{1}{n}{\parallel v\parallel }_X\ge 0$ . Replace v with −v. Then, we obtain $\langle -I_{\lambda }^{\prime },v\rangle +\frac{1}{n}{\parallel v\parallel }_X\ge 0$ . Thus, ${\parallel I_{\lambda }^{\prime }({\phi }_n)\parallel }_{X^{\ast }}\le \frac{1}{n}$ . We infer that there exists a sequence ${\{{\phi }_n\}}_{n\in \mathbb{N}}\subset B_{\rho }(0)$ such that

$I_{\lambda }({\phi }_n)\to \hat{c}<0,{\parallel I_{\lambda }^{\prime }({\phi }_n)\parallel }_{X^{\ast }}\to 0,n\to \mathrm{\infty }.$

Graph

By Lemma 4.2, there is ${\phi }_n\to {\phi }_0^{(2)}$ in X. Then, we have $I_{\lambda }^{\prime }({\phi }_0^{(2)})=0$ and $I_{\lambda }({\phi }_0^{(2)})=\hat{c}<0$ , that is, ${\phi }_0^{(2)}$ is another solution of equations (1) with negative energy, which ends the proof. □

Here, we give an example of application of Theorem 1.1.

Example 4.1

Let $\mathrm{\Omega }=\{(x,y)\in \mathbb{R}:x^2+y^2\le 1\}$ . Consider the problem

17 $\{\begin{array}{c}{(-\mathrm{\Delta })}_{x^2+y^2+3}^{\frac{1}{2}}\phi +{(-\mathrm{\Delta })}_{x^2+y^2+5}^{\frac{1}{2}}\phi +2|\phi {|}^{2x^2+1}\phi +4|\phi {|}^{2x^2+3}\phi \hfill \\ =|x||\phi {|}^{\frac{1}{6}}\phi +x^2|\phi {|}^{\frac{1}{2}}\phi ,x\in \mathrm{\Omega },\hfill \\ {\mathcal{N}}_{\frac{1}{2},x^2+y^2+3}\phi +{\mathcal{N}}_{\frac{1}{2},x^2+y^2+5}\phi +ln|x||\phi {|}^{2x^2+1}\phi \hfill \\ +ln|x||\phi {|}^{2x^2+3}\phi =0,x\in {\mathbb{R}}^N\setminus \bar{\mathrm{\Omega }}.\hfill \end{array}$

Graph

By simple calculations, we obtain $meas(\partial \mathrm{\Omega })=2\pi$ , $p_1^{-}=3$ , $p_1^{+}=4$ , $p_2^{-}=5$ , $p_2^{+}=6$ . Conditions (P), (G), (H), and (V) are satisfied. We observe that all assumptions of Theorem 1.1 are fulfilled. Hence, Theorem 1.1 implies that problem (17) admits two nontrivial weak solutions.

Proof of Theorem 1.2

To prove Theorem 1.2, we first recall the following lemmas.

Lemma 5.1

([[15]]) LetXbe a reflexive and separable Banach space. Then, there are $\{e_n\}\subset E$ and $\{e_n^{\ast }\}\subset E^{\ast }$ such that

$E=\overline{span\{e_n:n=1,2,3...\}},E^{\ast }=\overline{span\{e_n^{\ast }:n=1,2,3...\}}$

Graph

and

$\langle e_i^{\ast },e_j\rangle =\{\begin{array}{cc}1,\hfill & \mathit{\text{if}}i=j,\hfill \\ 0,\hfill & \mathit{\text{if}}i\ne j.\hfill \end{array}$

Graph

Denote

$E_n=span\{e_n\},X_k=\underset{n=1}{\overset{k}{⨁}}{E}_n,\mathit{\text{and}}{Y}_k=\stackrel{‾}{\underset{n=k}{\overset{\mathrm{\infty }}{⨁}}{E}_n}.$

Graph

Lemma 5.2

([[15]]) Assume that $q(x)\in C_{+}(\bar{\mathrm{\Omega }})$ , $q(x)<p^{\ast }(x)$ , for any $x\in \bar{\mathrm{\Omega }}$ and denote

${\tilde{\xi }}_k=\underset{\phi \in Y_k,{\parallel \phi \parallel }_X=1}{sup}{\parallel \phi \parallel }_{L^{q(x)}(\mathrm{\Omega })},$

Graph

then ${lim}_{k\to \mathrm{\infty }}{\tilde{\xi }}_k=0$ .

Now, we recall the fountain theorem.

Theorem 5.1

([[15]]) LetXbe a real Banach space and $I_{\lambda }\in C^1(X,\mathbb{R})$ be a even functional satisfying the $(PS)$ conditions. There exists $r_k>0$ such that ${\hat{\rho }}_k>r_k>0$ for every $k\in \mathbb{N}$ . Then, the following conditions hold:

• ${\alpha }_k=max\{I_{\lambda }(\phi ):\phi \in X_k,\parallel \phi \parallel ={\hat{\rho }}_k\}\le 0$ ;
• ${\beta }_k=inf\{I_{\lambda }(\phi ):\phi \in Y_k,\parallel \phi \parallel =r_k\}\to +\mathrm{\infty }$ as $k\to \mathrm{\infty }$ .

Then, $I_{\lambda }$ possesses a series of critical points ${\phi }_k$ such that $I_{\lambda }({\phi }_k)\to +\mathrm{\infty }$ .

Lemma 5.3

Assume that assumptions (P), (G), (V), and (H) hold. There exists ${\hat{\rho }}_k>0$ such that

$\underset{\phi \in X_k,\parallel \phi \parallel ={\hat{\rho }}_k}{max}{I}_{\lambda }(\phi )<0.$

Graph

Proof

Let $t\in (0,1)$ . For ${\parallel \hat{\phi }\parallel }_X={\tilde{\rho }}_k\ge 1$ and ${\hat{\rho }}_k>{\tilde{\rho }}_k$ , there exists φ̂ such that

$\begin{array}{rl}{I}_{\lambda }(t\hat{\phi })\le & t^{p_2^{+}}\sum _{i=1}^2({\int }_{{\mathbb{R}}^{2N}\setminus {(\mathcal{C}\mathrm{\Omega })}^2}\frac{|\hat{\phi }(x)-\hat{\phi }(y){|}^{p_i(x,y)}}{2p_i(x,y)|x-y{|}^{N+s{p}_i(x,y)}}dxdy+{\int }_{\mathrm{\Omega }}{V}_i(x)\frac{|\hat{\phi }{|}^{{\overline{p}}_i(x)}}{{\overline{p}}_i(x)}dx\\ & +{\int }_{{\mathbb{R}}^N\setminus \mathrm{\Omega }}\frac{\beta (x)|\hat{\phi }{|}^{{\overline{p}}_i(x)}}{{\overline{p}}_i(x)}dx)-{\lambda }_1{t}^{{r_1}^{-}}{\int }_{\mathrm{\Omega }}\frac{A_1(x)|\hat{\phi }{|}^{r_1(x)}}{r_1(x)}dx\\ & -{\lambda }_2{t}^{{r_2}^{-}}{\int }_{\mathrm{\Omega }}\frac{A_2(x)|\hat{\phi }{|}^{r_2(x)}}{r_2(x)}dx-{\int }_{{\mathbb{R}}^N\setminus \mathrm{\Omega }}t\hat{\phi }{g}_1(x)dx-{\int }_{{\mathbb{R}}^N\setminus \mathrm{\Omega }}t\hat{\phi }{g}_2(x)dx\\ \le & \frac{t^{p_2^{+}}}{p_1^{-}}{\parallel \hat{\phi }\parallel }_X^{p_2^{+}}-\frac{{\lambda }_1{t}^{{r_1}^{-}}}{r_1^{+}}{\int }_{\mathrm{\Omega }}{A}_1(x)|\hat{\phi }{|}^{r_1(x)}dx<0\end{array}$

Graph

with $p_1^{-}>r_1^{+}>1$ . Taking $\phi =t\hat{\phi }$ , for sufficiently small t, it follows that

${\alpha }_k=\underset{\phi \in X_k,\parallel \phi \parallel ={\hat{\rho }}_k}{max}{I}_{\lambda }(\phi )\le 0.$

Graph

□

Lemma 5.4

Assume that assumptions (P), (G), (V), and (H) hold. There exists $r_k>0$ such that

$\underset{\phi \in Y_k,\parallel \phi \parallel =r_k}{inf}{I}_{\lambda }(\phi )>+\mathrm{\infty }.$

Graph

Proof

According to Lemma 5.2, for ${\parallel \phi \parallel }_X=r_k>1$ , we obtain

$\begin{array}{rl}{I}_{\lambda }(\phi )\ge & \frac{1}{p_2^{+}}{\parallel \phi \parallel }_X^{p_1^{-}}-\frac{{\lambda }_1}{r_1^{-}}{\int }_{\mathrm{\Omega }}{A}_1(x)|\phi {|}^{r_1(x)}dx-\frac{{\lambda }_2}{r_2^{-}}{\int }_{\mathrm{\Omega }}{A}_2(x)|\phi {|}^{r_2(x)}dx\\ \ge & \frac{1}{p_2^{+}}{\parallel \phi \parallel }_X^{p_1^{-}}-\frac{2{\lambda }_1{M}_1^{r_1^{+}}}{r_1^{-}}{\parallel A_1\parallel }_{s_1(x)}{\parallel \phi \parallel }_X^{r_1^{+}}-\frac{2{\lambda }_2{M}_2^{r_2^{+}}}{r_2^{-}}{\parallel A_2\parallel }_{s_2(x)}{\parallel \phi \parallel }_X^{r_2^{+}}\\ \ge & \frac{1}{p_2^{+}}{\parallel \phi \parallel }_X^{p_1^{-}}-\frac{2{\lambda }_1{M}_1^{r_1^{+}}{\tilde{\xi }}_k^{r_1^{+}}}{r_1^{-}}{\parallel A_1\parallel }_{s_1(x)}{\parallel \phi \parallel }_X^{r_1^{+}}-\frac{2{\lambda }_2{M}_2^{r_2^{+}}{\tilde{\xi }}_k^{r_2^{+}}}{r_2^{-}}{\parallel A_2\parallel }_{s_2(x)}{\parallel \phi \parallel }_X^{r_2^{+}}.\end{array}$

Graph

Let

$\begin{array}{rl}& max\{\frac{2{\lambda }_1{M}_1^{r_1^{+}}{\tilde{\xi }}_k^{r_1^{+}}}{r_1^{-}}{\parallel A_1\parallel }_{s_1(x)}{\parallel \phi \parallel }_X^{r_1^{+}},\frac{2{\lambda }_2{M}_2^{r_2^{+}}{\tilde{\xi }}_k^{r_2^{+}}}{r_2^{-}}{\parallel A_2\parallel }_{s_2(x)}{\parallel \phi \parallel }_X^{r_2^{+}}\}\\ & =2\frac{{\lambda }_{\phi }{M}_{\phi }^{r_{\phi }}{\tilde{\xi }}_k^{\phi }}{r_{\phi }^{-}}{A}_{\phi }^{\ast }{\parallel \phi \parallel }_X^{r_{\phi }},\end{array}$

Graph

and there exists a constant C̃ such that $r_{\phi }=max\{r_1^{+}+\tilde{C},r_2^{+}+\tilde{C}\}$ , where $1<p_1^{-}<r_{\phi }^{-}<r_{\phi }$ . Therefore,

$\begin{array}{rl}& \frac{1}{p_2^{+}}{\parallel \phi \parallel }_X^{p_1^{-}}-\frac{2{\lambda }_1{M}_1^{r_1^{+}}{\tilde{\xi }}_k^{r_1^{+}}}{r_1^{-}}{\parallel A_1\parallel }_{s_1(x)}{\parallel \phi \parallel }_X^{r_1^{+}}-\frac{2{\lambda }_2{M}_2^{r_2^{+}}{\tilde{\xi }}_k^{r_2^{+}}}{r_2^{-}}{\parallel A_2\parallel }_{s_2(x)}{\parallel \phi \parallel }_X^{r_2^{+}}\\ & \ge {\parallel \phi \parallel }_X^{p_1^{-}}(\frac{1}{p_2^{+}}-4\frac{{\lambda }_{\phi }{M}_{\phi }^{r_{\phi }}{\tilde{\xi }}_k^{\phi }}{r_{\phi }^{-}}{A}_{\phi }^{\ast }{\parallel \phi \parallel }_X^{r_{\phi }-p_1^{-}}).\end{array}$

Graph

Choose

$r_k={\left(\frac{5{\lambda }_{\phi }{M}_{\phi }^{r_{\phi }}{p}_2^{+}{\tilde{\xi }}_k^{\phi }}{r_{\phi }^{-}}{A}_{\phi }^{\ast }\right)}^{\frac{1}{p_1^{-}-r_{\phi }}}.$

Graph

Since $p_1^{-}<r_{\phi }$ , we have $r_k\to +\mathrm{\infty }$ as $k\to +\mathrm{\infty }$ . By the choice of $r_k$ with ${\parallel \phi \parallel }_X=r_k$ such that ${\hat{\rho }}_k>{\tilde{\rho }}_k>r_k>0$ , we obtain

${\beta }_k=\underset{\phi \in Y_k,\parallel \phi \parallel =r_k}{inf}{I}_{\lambda }(\phi )\to +\mathrm{\infty },k\to +\mathrm{\infty }.$

Graph

□

Proof of Theorem 1.2

Let hypotheses (P), (G), (V), and (H) be satisfied. By Lemma 4.2, $I_{\lambda }$ satisfies $(PS)$ conditions. Under the definition of $I_{\lambda }$ in Lemma 3.13, it follows that $I_{\lambda }(0)=0$ and $I_{\lambda }$ is an even function. Therefore, from Lemmas 5.3 and 5.4, it can be deduced that $I_{\lambda }$ satisfies Theorem 5.1. Then, $I_{\lambda }$ possesses a series of critical points ${\phi }_k$ as $k\to +\mathrm{\infty }$ . In conclusion, equations (1) possess infinitely many nontrivial weak solutions. □

Here, we give an example of application of Theorem 1.2.

Example 5.1

Let $\mathrm{\Omega }=\{(x,y)\in \mathbb{R}:x^2+y^2\le 1\}$ . Consider the problem

18 $\{\begin{array}{c}{(-\mathrm{\Delta })}_{x^2+y^2+3}^{\frac{1}{2}}\phi +{(-\mathrm{\Delta })}_{x^2+y^2+5}^{\frac{1}{2}}\phi +(x^2+1)|\phi {|}^{2x^2+1}\phi +(e^{|x|}+2)|\phi {|}^{2x^2+3}\phi \hfill \\ =|x||\phi {|}^{\sqrt{|x|}-\frac{3}{4}}\phi +x^2|\phi {|}^{\sqrt{|x|}-\frac{1}{4}}\phi ,x\in \mathrm{\Omega },\hfill \\ {\mathcal{N}}_{\frac{1}{2},x^2+y^2+3}\phi +{\mathcal{N}}_{\frac{1}{2},x^2+y^2+5}\phi +ln|x||\phi {|}^{2x^2+1}\phi \hfill \\ +ln|x||\phi {|}^{2x^2+3}\phi =0,x\in {\mathbb{R}}^N\setminus \bar{\mathrm{\Omega }}.\hfill \end{array}$

Graph

By simple calculations, we obtain $meas(\partial \mathrm{\Omega })=2\pi$ , $p_1^{-}=3$ , $p_1^{+}=4$ , $p_2^{-}=5$ , $p_2^{+}=6$ , $r_1^{-}=5/4$ , $r_1^{+}=9/4$ , $r_2^{-}=7/4$ , and $r_2^{+}=11/4$ . That is, conditions (P), (G), (H), and (V) are satisfied. We observe that all assumptions of Theorem 1.2 are fulfilled. Hence, Theorem 1.2 implies that problem (18) admits infinitely many nontrivial weak solutions.

Proof of Theorem 1.3

We give some results with the aid of the Krasnoselskii genus. Let E be a real Banach space and set

$\mathfrak{R}=\{\mathcal{A}\subset E\setminus \{0\}:\mathcal{A}\text{is compact and}\mathcal{A}=-\mathcal{A}\}.$

Graph

Let $\mathcal{A}\subset \mathfrak{R}$ and $E={\mathbb{R}}^k$ . We define genus

$\gamma (\mathcal{A})=min\{k\ge 1:\text{there exists an odd continuous mapping}\psi :\mathcal{A}\to {\mathbb{R}}^k\setminus \{0\}\}.$

Graph

If the mapping ψ does not exist for any $k>0$ , and set $\gamma (\mathcal{A})=\mathrm{\infty }$ . If $\mathcal{A}$ is a subset consisting of a finite number of pairs of points, then, $\gamma (\mathcal{A})=1$ . Furthermore, from definition, $\gamma (\mathrm{\varnothing })=0$ .

Lemma 6.1

([[18]]) Let $E=R^N$ and∂Ω be the boundary of an open, symmetric, and bounded subset $\partial \mathrm{\Omega }\subset R^N$ with $0\in \mathrm{\Omega }$ . Then, $\gamma (\partial \mathrm{\Omega })=N$ .

Corollary 6.1

([[18]]) $\gamma (S^{N-1})=N$ .

Theorem 6.1

([[18]]) Let $I_{\lambda }\in C^1(X)$ be a functional satisfying the $(PS)$ conditions and assume that

• $I_{\lambda }$ is bounded from below and even;
• there is a compact set $K\in \mathbb{R}$ such that $\gamma (K)=k$ and ${sup}_{x\in K}{I}_{\lambda }(x)<I_{\lambda }(0)$ .

Then, $I_{\lambda }$ has at leastkpairs of distinct critical points whose corresponding critical values are all less than $I_{\lambda }(0)$ .

Proof of Theorem 1.3

Combining (12), (13), and the Hölder inequality (7) for ${\parallel \phi \parallel }_X>1$ , we obtain

$\begin{array}{rl}{I}_{\lambda }(\phi )\ge & \frac{1}{p_2^{+}}{\parallel \phi \parallel }_X^{p_1^{-}}-\frac{{\lambda }_1}{r_1^{-}}{\parallel A_1\parallel }_{s_1(x)}{\parallel \phi \parallel }_{L^{s_1^{\prime }(x)r_1(x)}(\mathrm{\Omega })}^{r_1^{+}}-\frac{{\lambda }_2}{r_2^{-}}{\parallel A_2\parallel }_{s_2(x)}{\parallel \phi \parallel }_{L^{s_2^{\prime }(x)r_2(x)}(\mathrm{\Omega })}^{r_2^{+}}\\ \ge & \frac{1}{p_2^{+}}{\parallel \phi \parallel }_X^{p_1^{-}}-\frac{{\lambda }_1}{r_1^{-}}2{\parallel A_1\parallel }_{s_1(x)}{M}_1^{r_1^{+}}{\parallel \phi \parallel }_X^{r_1^{+}}-\frac{{\lambda }_2}{r_2^{-}}2{\parallel A_2\parallel }_{s_2(x)}{M}_2^{r_2^{+}}{\parallel \phi \parallel }_X^{r_2^{+}}.\end{array}$

Graph

Since $max\{1,r_1^{+},r_2^{+}\}<p_1^{-}$ , for ${\parallel \phi \parallel }_X$ large enough, $I_{\lambda }$ is bounded from below. $I_{\lambda }\in X$ is an even function by the definition and $I_{\lambda }(0)=0$ . Moreover, $I_{\lambda }$ is coercive in X and satisfies the $(PS)$ conditions by Lemma 4.2. Let

${\mathfrak{R}}_k=\{\mathcal{M}\subset \mathfrak{R}:\gamma (\mathcal{M})\ge k\},c_k=\underset{\mathcal{M}\in {\mathfrak{R}}_k}{inf}\underset{\phi \in \mathcal{M}}{sup}I(\phi ),k=1,2,....$

Graph

We obtain

$-\mathrm{\infty }<c_1\le c_2\le \cdots \le c_k\le c_{k+1}\le \cdot \cdot \cdot .$

Graph

Now we prove that for any $k\in \mathbb{N}$ , there is $c_k<0$ . For each k, we take k disjoint open sets ${\tilde{K}}_i$ such that ${\bigcup }_{i=1}^k{\tilde{K}}_i\subset \mathrm{\Omega }$ . For $i=1,...,k$ , let ${\phi }_i\in (X\bigcap C_0^{\mathrm{\infty }}({\tilde{K}}_i)\setminus \{0\}$ with ${\parallel {\phi }_i\parallel }_X=1$ , and

${\mathcal{M}}_k=span\{{\phi }_1,{\phi }_2,...,{\phi }_k\}.$

Graph

Since each norm on ${\mathcal{M}}_k$ is equivalent, there is ${\rho }_k^{\ast }\in (0,1)$ such that $\phi \in {\mathcal{M}}_k$ with ${\parallel \phi \parallel }_X\le {\rho }_k^{\ast }$ , which means that ${\parallel \phi \parallel }_{\mathrm{\infty }}<C_{{\rho }_k^{\ast }}<1$ . Set

$S_{{\rho }_k^{\ast }}^{(k)}=\{\phi \in {\mathcal{M}}_k:{\parallel \phi \parallel }_X={\rho }_k^{\ast }\}.$

Graph

Combining the compactness of $S_{{\rho }_k^{\ast }}^{(k)}$ and $t\in (0,1)$ for all $\phi \in S_{{\rho }_k^{\ast }}^{(k)}$ ,

$\begin{array}{rl}{I}_{\lambda }(t\phi )\le & t^{p_1^{-}}\sum _{i=1}^2({\int }_{{\mathbb{R}}^{2N}\setminus {(\mathcal{C}\mathrm{\Omega })}^2}\frac{|\phi (x)-\phi (y){|}^{p_i(x,y)}}{2p_i(x,y)|x-y{|}^{N+s{p}_i(x,y)}}dxdy+{\int }_{\mathrm{\Omega }}{V}_i(x)\frac{|\phi {|}^{{\overline{p}}_i(x)}}{{\overline{p}}_i(x)}dx\\ & +{\int }_{{\mathbb{R}}^N\setminus \mathrm{\Omega }}\frac{\beta (x)|\phi {|}^{{\overline{p}}_i(x)}}{{\overline{p}}_i(x)}dx)-{\int }_{{\mathbb{R}}^N\setminus \mathrm{\Omega }}t\phi g_1(x)dx-{\int }_{{\mathbb{R}}^N\setminus \mathrm{\Omega }}t\phi g_2(x)dx\\ & -{\lambda }_1{t}^{r_1^{-}}{\int }_{\mathrm{\Omega }}\frac{A_1(x)|\phi {|}^{r_1(x)}}{r_1(x)}dx-{\lambda }_2{t}^{r_2^{-}}{\int }_{\mathrm{\Omega }}\frac{A_2(x)|\phi {|}^{r_2(x)}}{r_2(x)}dx\\ \le & 3K_1{t}^{p_1^{-}}{\parallel \phi \parallel }_X^{p_2^{+}}-\frac{{\lambda }_1{t}^{r_1^{-}}}{p_1^{-}}{\int }_{\mathrm{\Omega }}{A}_1(x)|\phi {|}^{r_1(x)}dx.\end{array}$

Graph

Sine $1<r_1^{-}<p_1^{-}$ , there exist $t_k\in (0,1)$ and ${\epsilon }_k$ such that

$I_{\lambda }(t_k\phi )<-{\epsilon }_k<0.$

Graph

Thus, $I_{\lambda }(\phi )<0$ for all $\phi \in S_{t_k{\rho }_k^{\ast }}^{(k)}$ . Furthermore, $\gamma (S_{t_k{\rho }_k^{\ast }}^{(k)})=k$ such that $c_k<-{\epsilon }_k<0$ for all k, and the assertion is proved. Each $c_k$ is a critical value by the Krasnoselskii genus theory. Combining Theorem 6.1, $I_{\lambda }$ has at least k pairs of different critical points. In addition, since k is arbitrary, we obtain an infinite number of critical points of equations (1). □

Acknowledgements

This work is supported by the Postgraduate Research Practice Innovation Program of Jiangsu Province under Grant (KYCX23-0669); and the Doctoral Foundation of Fuyang Normal University (2023KYQD0044).

Author contributions

Zhenfeng Zhang wrote the main manuscript text and Tianqing An, Weichun Bu, and Shuai Li verified this article. All authors reviewed the manuscript.

Funding

This work is supported by the Postgraduate Research Practice Innovation Program of Jiangsu Province under Grant (KYCX23-0669); and the Doctoral Foundation of Fuyang Normal University (2023KYQD0044).

Data Availability

No datasets were generated or analysed during the current study.

Declarations

Competing interests

The authors declare no competing interests.

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

1 Benci V, Fortunato D, Pisani L. Soliton like solutions of a Lorentz invariant equation in dimension 3. Rev. Math. Phys. 1998; 10; 3: 315-344. 1626832. 10.1142/S0129055X98000100 2 Jia G, Zhang L. Multiplicity of solutions for singular semilinear elliptic equations in weighted Sobolev spaces. Bound. Value Probl. 2014; 2014. 3286094. 10.1186/s13661-014-0156-9 3 Mihăilescu M. On a class of nonlinear problems involving a -Laplace type operator. Czechoslov. Math. J. 2008; 58: 155-172. 2402532. 10.1007/s10587-008-0011-1 4 Chung N.T, Toan H.Q. On a class of fractional Laplacian problems with variable exponents and indefinite weights. Collect. Math. 2020; 71: 223-237. 4083643. 10.1007/s13348-019-00254-5 5 Ghobadi A, Heidarkhani S. Multiple solutions for nonlocal fractional Kirchhoff type problems. Differ. Equ. Appl. 2022; 14; 4: 597-608. 4532717 6 Gharehgazlouei F, Graef J.R, Heidarkhani S, Kong L. Existence and multiplicity of solutions to a fractional p-Laplacian elliptic Dirichlet problem. Electron. J. Differ. Equ. 2023; 2023. 4615681 7 Liao F.F, Heidarkhani S, Salari A. Existence results for Kirchhoff nonlocal fractional equations. Kragujev. J. Math. 2025; 49; 1: 17-30 8 Heidarkhani S, De Araujo A.L.A, Afrouzi G.A, Moradi S. Multiple solutions for Kirchhoff type problems with variable exponent and nonhomogeneous Neumann conditions. Math. Nachr. 2018; 291; 2–3: 326-342. 3767141. 10.1002/mana.201600425 9 Heidarkhani S, Ghobadi A, Avci M. Multiple solutions for a class of -Kirchhoff type equations. Appl. Math. E-Notes. 2022; 22: 160-168. 4438423 Heidarkhani S, Moradi S, Afrouzi G.A. Existence of one weak solution for a Steklov problem involving the weighted -Laplacian. J. Nonlinear Funct. Anal. 2023; 2023; 8: 1-15 Zuo J, Fiscella A, Bahrouni A. Existence and multiplicity results for fractional Choquard problems with variable order. Complex Var. Elliptic Equ. 2022; 67: 500-516. 4386507. 10.1080/17476933.2020.1835878 Biswas R, Bahrouni S, Carvalho M.L. Fractional double phase Robin problem involving variable order-exponents without Ambrosetti-Rabinowitz condition. Z. Angew. Math. Phys. 2022; 73; 99: 1-24. 4411072 Laskin N. Fractional quantum mechanics and Lévy path integrals. Phys. Lett. A. 2000; 268: 298-305. 2000PhLA.268.298L. 1755089. 1:CAS:528:DC%2BD3cXislChtb0%3D. 10.1016/S0375-9601(00)00201-2 Xiang M, Zhang B, Yang D. Multiplicity results for variable-order fractional Laplacian equations with variable growth. Nonlinear Anal. 2019; 178: 190-204. 3886611. 10.1016/j.na.2018.07.016 Bu W, An T, Sousa J, da Vanteler C, Yun Y. Infinitely many solutions for fractional p-Laplacian Schrödinger-Kirchhoff type equations with symmetric variable-order. Symmetry. 2021; 13: 1-22. 10.3390/sym13081393 Ambrosetti A, Brézis H, Cerami G. Combined effects of concave and convex nonlinearities in some elliptic problems. J. Funct. Anal. 1994; 122; 2: 519-543. 1276168. 10.1006/jfan.1994.1078 Ho K, Sim I. Existence and multiplicity of solutions for degenerate -Laplace equations involving concave-convex type nonlinearities with two parameters. Taiwan. J. Math. 2015; 19; 5: 1469-1493. 3412016. 10.11650/tjm.19.2015.5187 Pucci P, Xiang M, Zhang B. Existence and multiplicity of entire solutions for fractional p-Kirchhoff equations. Adv. Nonlinear Anal. 2015; 5: 27-55. 3456737. 10.1515/anona-2015-0102 Kim J.M, Kim Y.H. Multiple solutions to the double phase problems involving concave-convex nonlinearities. AIMS Math. 2023; 8: 5060-5079. 4525837. 10.3934/math.2023254 Jebelean P, Serban C. Relativistic type systems with parametric odd nonlinearities. Commun. Pure Appl. Anal. 2023; 22; 3: 736-750. 4550197. 10.3934/cpaa.2023007 Mugnai D, Perera K, Lippi E.P. A priori estimates for the fractional p-Laplacian with nonlocal Neumann boundary conditions and applications. Commun. Pure Appl. Anal. 2022; 21; 11: 275-292. 4507380 Deng S. Positive solutions for Robin problem involving the -Laplacian. J. Math. Anal. Appl. 2009; 360: 548-560. 2561253. 10.1016/j.jmaa.2009.06.032 Manouni S.E, Marino G, Winkert P. Existence results for double phase problems depending on Robin and Steklov eigenvalues for the p-Laplacian. Adv. Nonlinear Anal. 2022; 11: 304-320. 4291979. 10.1515/anona-2020-0193 Perera K, Pucci P, Varga C. An existence result for a class of quasilinear elliptic eigenvalue problems in unbounded domains. NoDEA Nonlinear Differ. Equ. Appl. 2014; 21: 441-451. 3211040. 10.1007/s00030-013-0255-9 Fragnelli G, Mugnai D, Papageorgiou N.S. Positive and nodal solutions for parametric nonlinear Robin problems with indefinite potential. Discrete Contin. Dyn. Syst. 2016; 36; 11: 6133-6166. 3543584. 10.3934/dcds.2016068 Jebelean P, Serban C. Boundary value problems for discontinuous perturbations of singular ϕ-Laplacian operator. J. Math. Anal. Appl. 2015; 431: 662-681. 3357606. 10.1016/j.jmaa.2015.06.004 Eddine N.C, Ragusa M.A. Generalized critical Kirchhoff-type potential systems with Neumann boundary conditions. Appl. Anal. 2022; 101: 3958-3988. 4462176. 10.1080/00036811.2022.2057305 Bahrouni A, Radulescu V.D, Winkert P. Robin fractional problems with symmetric variable growth. J. Math. Phys. 2020; 61: 1-14. 4157423. 10.1063/5.0014915 Irzi N, Kefi K. The fractional -Neumann boundary conditions for the nonlocal -Laplacian operator. Appl. Anal. 2021; 102: 839-851. 4568523. 10.1080/00036811.2021.1965585 Allalou C, Ouaarabi M.E, Melliani S. Existence and uniqueness results for a class of -Kirchhoff-type problems with convection term and Neumann boundary data. J. Elliptic Parabolic Equ. 2022; 8: 617-633. 4426351. 10.1007/s41808-022-00165-w Fan X, Zhao D. On the spaces and. J. Math. Anal. Appl. 2001; 263; 2: 424-446. 1866056. 10.1006/jmaa.2000.7617 Mashiyev R.A, Cekic B, Avci M, Yucedag Z. Existence and multiplicity of weak solutions for nonuniformly elliptic equations with nonstandard growth condition. Complex Var. Elliptic Equ. 2012; 57; 5: 579-595. 2903484. 10.1080/17476933.2011.598928 Edmunds D.E, Rakosnik J. Sobolev embeddings with variable exponent. Stud. Math. 2000; 143; 3: 267-293. 1815935. 10.4064/sm-143-3-267-293 Bahrouni A, Rădulescu V.D. On a new fractional Sobolev space and applications to nonlocal variational problems with variable exponent. Discrete Contin. Dyn. Syst, Ser. S. 2018; 11; 3: 379-389. 3732173 Kim I.H, Kim Y.H, Park K. Existence and multiplicity of solutions for Schrödinger-Kirchhoff type problems involving the fractional -Laplacian in. Bound. Value Probl. 2020; 121: 1-24 Pezzo L.M.D, Rossi J.D. Traces for fractional Sobolev spaces with variable exponents. Adv. Oper. Theory. 2017; 2: 435-446. 3730039 Guariglia E. Riemann zeta fractional derivative functional equation and link with primes. Adv. Differ. Equ. 2019; 2019. 3978987. 10.1186/s13662-019-2202-5

By Zhenfeng Zhang; Tianqing An; Weichun Bu and Shuai Li

Reported by Author; Author; Author; Author