Home Regularity for critical fractional Choquard equation with singular potential and its applications
Article Open Access

Regularity for critical fractional Choquard equation with singular potential and its applications

  • Senli Liu , Jie Yang and Yu Su EMAIL logo
Published/Copyright: May 9, 2024

Abstract

We study the following fractional Choquard equation

( Δ ) s u + u x θ = ( I α * F ( u ) ) f ( u ) , x R N ,

where N 3 , s 1 2 , 1 , α ( 0 , N ) , θ ( 0 , 2 s ) , and I α is the Riesz potential. The main purpose of this article is twofold. We first study the regularity of weak solutions for the aforementioned equation with critical nonlinearity, which extends the results of θ = 0 in Moroz-Van Schaftingen [Existence of groundstates for a class of nonlinear Choquardequations, Trans. Amer. Math. Soc. 367 (2015), no. 9, 6557–6579]. Then, as an application of the regularity results, we establish the existence of ground state solutions for above equation with the nonlinearity involving embedding top and bottom indices, which is related to the Hardy-Littlewood-Sobolev inequality and singular term 1 x θ . It is worth noting that our approach is not involving the concentration-compactness principle.

MSC 2010: 35B25; 35J20; 35J75; 35Q55; 47J30

1 Introduction and main results

In this article, we focus our attention on the following equation:

(1.1) ( Δ ) s u + u x θ = ( I α * F ( u ) ) f ( u ) , x R N ,

where N 3 , s 1 2 , 1 , α ( 0 , N ) , and θ ( 0 , 2 s ) . The Riesz potential I α is defined by

I α ( x ) = Γ N α 2 2 α π N 2 Γ N 2 x N α , x R N \ { 0 } .

( Δ ) s is the fractional Laplacian, which can be expressed by

( Δ ) s u ( x ) = C N , s P.V. R N u ( x ) u ( y ) x y N + 2 s d y = C N , s lim ε 0 R N \ B ε ( x ) u ( x ) u ( y ) x y N + 2 s d y ,

where C N , s = π 2 s + N 2 Γ N 2 + s Γ ( s ) . Concerning for fractional Laplacian, see the previous studies [18,23] for more details.

Equation (1.1) comes from the research of solitary wave solution for the focusing time-dependent Hartree equation:

(1.2) i ψ t = ( Δ ) s ψ + 1 x θ ξ ψ ( I α * F ( ψ ) ) f ( ψ ) , ( t , x ) R × R N ,

where i is the imaginary unit and ξ R . As is well known, the solitary wave solution of equation (1.2) has the form ψ ( t , x ) = e i ξ t u ( x ) , where u ( x ) solves equation (1.1). Equation (1.2) originates from many fields of physics, such as fractional quantum mechanics [12], Lévy processes [22], one component plasma [24], and cosmology [7]. For more details to the physical background, we refer the readers to the previous studies [19,27,29,34] and the references therein.

Mathematically, when s = 1 , θ = 0 , and F ( u ) = u 2 2 , equation (1.1) turns into

(1.3) Δ u + u = 1 2 ( I α * u 2 ) u , x R N ,

which was introduced by Pekar [35]. The early results of existence and symmetry for (1.3) are credited to Lieb and Lions. Lieb [24] established the existence and uniqueness, up to translations, of ground state solutions. Lions [26] studied the existence of a sequence of radially symmetric normalized solutions. Subsequently, equation (1.1) with s = 1 and θ = 0 has been widely studied under various assumptions on the nonlinearity f in past decades. The existence of ground state solutions, semi-classical state solutions, sign-changing solutions, bound state solutions, and multi-bump solutions were investigated in the previous studies [1,13,15,28,30,31,45]. The qualitative properties such as positivity, regularity, radial symmetry, and decay property of solutions were established in the previous studies [10,14,32,36,41,46].

When s 1 and θ = 0 , D’Avenia et al. [17] considered equation (1.1) with f ( u ) = u p 2 u . The authors showed the existence of ground state solutions of such equation if p N + α N , N + α N 2 s . In addition, the regularity, symmetry, as well as decays property of these solutions are also established. He and Rădulescu [21] investigated some existence results for the following equation:

(1.4) ( Δ ) s u + V ( x ) u = ( I α * u 2 α , s * ) u 2 α , s * 2 u , x R N , u D s , 2 ( R N ) , u ( x ) > 0 , x R N ,

where s ( 0 , 1 ) , N > 2 s , and α ( 0 , min { N , 4 s } ) . Under some suitable assumptions on V ( x ) , the authors established a version of nonlocal global compactness lemma for fractional Choquard equation with critical growth. As a consequence, they obtained some compactness results and the existence of positive solution to equation (1.4). For more recent works about fractional Choquard equation, see the previous studies [2,5,9] and the references therein.

The study of problems involving singular potential have a long history, some first results can be found in the previous studies [3,4,16,39,40]. Recently, Feng and Su [20] extended the results of Badiale and Rolando [3] to a fractional equation as follows:

(1.5) ( Δ ) s u + V ( x ) u = u 2 s , θ * 2 u + λ u q 2 u + u 2 s * 2 u , x R N ,

where N 3 , s 1 2 , 1 , θ ( 0 , 2 s ) , 2 s , θ * 2 + 2 θ N 2 s , 2 s * 2 N N 2 s , and V is a singular potential, which satisfies

  1. there exists an A > 0 such that V ( t ) A t θ for almost every t > 0 ,

  2. V L 1 ( a , b ) for some ( a , b ) with b > a > 0 .

By virtue of [38, Compactness lemma] and Lemma 2.1, the authors established the following continuous and compactness embeddings results, which are essential to study the functional framework by variational methods.

Proposition 1.1

[20] Assume that N 3 , s 1 2 , 1 and conditions ( V 1 ) ( V 2 ) are satisfied. Then the following continuous embeddings hold

W rad s , 2 ( R N , V ) L r ( R N ) , r [ 2 s , θ * , 2 s * ] , θ ( 0 , 2 s ) , W rad s , 2 ( R N , V ) L r ( R N ) , r [ 2 s * , 2 s , θ * ] , θ ( 2 s , 2 N 2 s ) .

Furthermore, the embeddings are compact if r 2 s , θ * and r 2 s * . D rad s , 2 ( R N ) is a radial subspace of D s , 2 ( R N ) . Moreover, D s , 2 ( R N ) and W rad s , 2 ( R N , V ) are defined by

D s , 2 ( R N ) u L 2 s * ( R N ) : R 2 N u ( x ) u ( y ) 2 x y N + 2 s d x d y <

and

W rad s , 2 ( R N , V ) u D rad s , 2 ( R N ) : u L 2 ( R N , V ) 2 = R N V ( x ) u 2 d x < .

To overcome the noncompactness embedding problems, Feng and Su [20] first proposed a compactness lemma, which can be regarded as a contrapositive proposition of the famous Lions vanishing lemma [42].

Proposition 1.2

[20] Assume that N 3 , s 1 2 , 1 , θ ( 0 , 2 s ) and conditions ( V 1 ) ( V 2 ) are satisfied. Let { u n } W rad s , 2 ( R N , V ) be any bounded sequence satisfying

lim n R N u n 2 s , θ * d x > 0 a n d lim n R N u n 2 s * d x > 0 .

Then the sequence { u n } converges weakly and a.e. to u 0 in L loc 2 ( R N ) .

By applying Propositions 1.1 and 1.2 and the Nehari manifold method, the authors obtained the existence of ground state solution for equation (1.5).

Inspired by the aforementioned works, especially t-th study by Feng and Su [20], we find a problem left whether the fractional Choquard equation involving the embedding top and bottom indices and a singular potential possesses nontrivial solutions. Further, can we obtain some regularity results of these solutions? To respond these questions, we first introduce some preliminaries, which will be of use.

The following well-known Hardy-Littlewood-Sobolev inequality is crucial to handle our problem.

Proposition 1.3

[25] Let r , t > 1 and α ( 0 , N ) with 1 r + 1 t = 1 + α N . Then, there exists C ( α , r , t ) > 0 such that for any u L r ( R N ) and v L t ( R N )

R 2 N u ( x ) v ( y ) x y N α d x d y C ( α , r , t ) u L r ( R N ) v L t ( R N ) .

If r = t = 2 N N + α , then C ( α , r , t ) = C N , α = π N α 2 Γ α 2 Γ N + α 2 Γ N 2 Γ ( N ) α N .

We handle equation (1.1) in the space

W rad s , 2 ( R N , θ ) u D rad s , 2 ( R N ) : u L 2 ( R N , θ ) 2 = R N u 2 x θ d x <

endowed with the norm

u W rad s , 2 ( R N , θ ) 2 u D s , 2 ( R N ) 2 + u L 2 ( R N , θ ) 2 = R 2 N u ( x ) u ( y ) 2 x y N + 2 s d x d y + R N u 2 x θ d x .

Equation (1.1) is variational, the energy functional can be defined by

Q ( u ) 1 2 u W rad s , 2 ( R N , θ ) 2 1 2 R N ( I α * F ( u ) ) F ( u ) d x .

It follows from Propositions 1.1 and 1.3 that Q is well defined. Moreover, Q is of class C 1 with the derivative given by

Q ( u ) , φ = R 2 N ( u ( x ) u ( y ) ) ( φ ( x ) φ ( y ) ) x y N + 2 s d x d y + R N u φ x θ d x R N [ I α * F ( u ) ] f ( u ) d x , φ W rad s , 2 ( R N , θ ) .

We say u is a weak solution of equation (1.1) if and only if Q ( u ) , φ = 0 for all φ W rad s , 2 ( R N , θ ) .

Usually, we know it is very tough to study the regularity of weak solutions. Furthermore, up to our knowledge, no results have been reported regarding the regularity of weak solution to equation (1.1). In this study, our goal is to fill this gap. The first result is based on the following assumtion:

  1. f C ( R , R ) and there exists C > 0 such that for every t R , t f ( t ) C ( t 2 s , α , θ + t 2 s , α * ) , where 2 s , α , θ ( N + α ) ( N + θ 2 s ) N ( N 2 s ) and 2 s , α * N + α N 2 s .

In this direction, our first main result can be stated as follows.

Theorem 1.1

Let s 1 2 , 1 , α N ( N 2 s ) N + 2 ( s θ ) , N ( N 2 s ) 2 θ N ( N θ ) ( N 2 s + θ ) ( N θ ) , condition ( F 1 ) hold and

(1.6) N > 4 s , f o r θ ( 0 , θ ˜ ) , N > N * , f o r θ θ ˜ , 2 s 3 ,

where N * = 2 ( θ s ) ( θ + 2 s ) + 2 θ ( 2 s θ ) 3 2 s 3 θ and θ ˜ = 8 82 6 159 3 82 + 6 159 3 3 s . If u W rad s , 2 ( R N , θ ) is a nontrivial solution of equation (1.1), then

  1. u L p ( R N ) for any p [ 2 s , α * , ] ,

  2. u satisfies the following Pohožaev identity

    P ( u ) = N 2 s 2 u D s , 2 ( R N ) 2 + N θ 2 u L 2 ( R N , θ ) 2 N + α 2 R N ( I α * F ( u ) ) F ( u ) d x = 0 .

Remark 1.1

For equation (1.1) with s = 1 and θ = 0 , Moroz and Van Schaftingen [33] obtain the nonlocal Brézis-Kato’s type regularity estimate and some regularity results of solutions. In the present article, we extend the results in [33] to s 1 2 , 1 and θ ( 0 , 2 s ) . For parameters s , θ , the rest cases for s 0 , 1 2 and θ ( 2 s , ) are still open. These problems are more challenging and interesting, and we are still working on them.

To prove Theorem 1.1, following the ideas of [33], we first establish the nonlocal Brézis-Kato’s type regularity estimate with singular potential (Lemma 2.4). With the help of radial inequality (Lemma 2.1), the Hardy inequality [37], and Lemma 2.4, we increase the integrability of the weak solution of equation (1.1). Subsequently, by means of a new Moser iteration method, we prove that any weak solution u of equation (1.1) satisfies u L ( R N ) . Finally, we conclude the Pohožaev identity of equation (1.1) by using some regularity results obtained in [44]. In particular, we believe that the methods used in this article can be adapted to Laplacian (that is s = 1 ), which are also new to the literature.

The second aim of this article is to give an application of Theorem 1.1. For this purpose, we study the following fractional Choquard equation:

(C) ( Δ ) s u + u x θ = I α * u 2 s , α , θ 2 s , α , θ + λ u q q + u 2 s , α * 2 s , α * ( u 2 s , α , θ 2 u + λ u q 2 u + u 2 s , α * 2 u ) , x R N .

Our second main result can be stated as follows.

Theorem 1.2

Let s 1 2 , 1 , α N ( N 2 s ) N + 2 ( s θ ) , N ( N 2 s ) 2 θ N ( N θ ) ( N 2 s + θ ) ( N θ ) and (1.6) hold. Then there exists Λ > 0 such that for any λ > Λ , equation ( C ) admits a nonnegative ground state solution.

Remark 1.2

When θ = 0 , Moroz and Van Schaftingen [33, p. 3, line 5] studied the generalized version of Berestycki and Lions condition [8] in sense of the Hartree-type nonlinearity, where the nonlinearity f C ( R , R ) satisfies the following conditions:

  1. There exists t 0 R \ { 0 } such that F ( t 0 ) 0 , where F : t R 0 t f ( ζ ) d ζ .

  2. There exists C > 0 such that for every t R , t f ( t ) C ( t N + α N + t N + α N 2 ) .

  3. F is subcritical:

    lim t 0 F ( t ) t N + α N = 0 and lim t F ( t ) t N + α N 2 = 0 .

Obviously, condition ( F 4 ) is not sharp, and it can be improved as follows:
  1. F is critical and there exist a 1 , a 2 > 0 such that

    lim t 0 F ( t ) t N + α N = a 1 and lim t F ( t ) t N + α N 2 = a 2 .

Let r N + α N , N + α N 2 . It follows from the Young inequality that

u r ( N 2 ) N r ( N 2 ) ( N + α ) 2 ( N + α ) u N + α N + N ( N + α ) ( N 2 ) N r 2 ( N + α ) u N + α N 2 .

As mentioned earlier, we know that conditions ( F 2 ) , ( F 3 ) , and ( F 5 ) are the “almost optimal” choice for the nonlinearity, which contains the embedding bottom and top indices.

Moreover, for θ ( 0 , 2 s ) , we can revise ( F 3 ) to ( F 1 ) and ( F 5 ) to the following condition:

  1. F is critical and there exist b 1 , b 2 > 0 such that

    lim t 0 F ( t ) t 2 s , α , θ = b 1 and lim t F ( t ) t 2 s , α * = b 2 , θ ( 0 , 2 s ) .

As one can see, the nonlinearity f in equation ( C ) satisfies conditions ( F 1 ) , ( F 2 ) , and ( F 6 ) . Therefore, we can affirm that this choice of f is “almost optimal.”

Remark 1.3

It follows from Proposition 1.3 that

R N ( I α * u 2 s , α , θ ) u 2 s , α , θ d x C ( N , α ) u L 2 s , θ * ( R N ) 2 2 s , α , θ

and

R N ( I α * u 2 s , α * ) u 2 s , α * d x C ( N , α ) u L 2 s * ( R N ) 2 2 s , α * .

In view of Proposition 1.1, the compactness of the following embeddings have disappeared

W rad s , 2 ( R N , θ ) L 2 s , θ * ( R N ) and W rad s , 2 ( R N , θ ) L 2 s * ( R N ) , θ ( 0 , 2 s ) ,

which set an obstacle to verify the compactness of the corresponding functional. Thus, we call 2 s , α , θ = ( N + α ) ( N + θ 2 s ) N ( N 2 s ) is the bottom embedding index and 2 s , α * = N + α N 2 s is the top embedding index with respect to the Hardy-Littlewood-Sobolev inequality and singular term 1 x θ .

We outline our strategy in the proof of Theorem 1.2. By Propositions 1.1 and 1.3, the corresponding functional J : W rad s , 2 ( R N , θ ) R to equation ( C ) can be defined by

J ( u ) 1 2 u W rad s , 2 ( R N , θ ) 2 1 2 ( 2 s , α , θ ) 2 R N ( I α * u 2 s , α , θ ) u 2 s , α , θ d x λ 2 2 q 2 R N ( I α * u q ) u q d x 1 2 ( 2 s , α * ) 2 R N ( I α * u 2 s , α * ) u 2 s , α * d x λ 2 s , α , θ q R N ( I α * u 2 s , α , θ ) u q d x λ q 2 s , α * R N ( I α * u q ) u 2 s , α * d x 1 2 s , α , θ 2 s , α * R N ( I α * u 2 s , α , θ ) u 2 s , α * d x .

Obviously, J is of class C 1 and well defined. Moreover, for φ W rad s , 2 ( R N , θ ) , we have

J ( u ) , φ = R 2 N ( u ( x ) u ( y ) ) ( φ ( x ) φ ( y ) ) x y N + 2 s d x d y + R N u φ x θ d x 1 2 s , α , θ R N ( I α * u 2 s , α , θ ) u 2 s , α , θ 2 u φ d x λ 2 q R N ( I α * u q ) u q 2 u φ d x 1 2 s , α * R N ( I α * u 2 s , α * ) u 2 s , α * 2 u φ d x λ q 2 s , α , θ + 2 s , α , θ q R N ( I α * u 2 s , α , θ ) u q 2 u φ d x λ 2 s , α * q + q 2 s , α * R N ( I α * u q ) u 2 s , α * 2 u φ d x 2 s , α * 2 s , α , θ + 2 s , α , θ 2 s , α * R N ( I α * u 2 s , α , θ ) u 2 s , α * 2 u φ d x .

To overcome the lack of compactness caused by the embedding top and bottom indicies, we consider the following subcritical perturbed equation:

( Δ ) s u + u x θ = I α * u 2 s , α , θ + ε 2 s , α , θ + ε + λ u q q + u 2 s , α * ε 2 s , α * ε ( C ε ) × ( u 2 s , α , θ + ε 2 u + λ u q + u 2 s , α * ε 2 u ) , x R N ,

where ε ( 0 , min { q 2 s , α , θ , 2 s , α * q } ) . The functional of equation ( C ε ) can be given by

J ε ( u ) 1 2 u W rad s , 2 ( R N , θ ) 2 1 2 ( 2 s , α , θ + ε ) 2 R N ( I α * u 2 s , α , θ + ε ) u 2 s , α , θ + ε d x λ 2 2 q 2 R N ( I α * u q ) u q d x 1 2 ( 2 s , α * ε ) 2 R N ( I α * u 2 s , α * ε ) u 2 s , α * ε d x λ ( 2 s , α , θ + ε ) q R N ( I α * u 2 s , α , θ + ε ) u q d x λ q ( 2 s , α * ε ) R N ( I α * u q ) u 2 s , α * ε d x 1 ( 2 s , α , θ + ε ) ( 2 s , α * ε ) R N ( I α * u 2 s , α , θ + ε ) u 2 s , α * ε d x .

By using the fact of Theorem 1.1, we know that all the weak solutions of equation ( C ε ) are classical solutions, and satisfy the following Pohožaev identity:

P ε ( u ) = N 2 s 2 u D s , 2 ( R N ) 2 + N θ 2 u L 2 ( R N , θ ) 2 N + α 2 ( 2 s , α , θ + ε ) 2 R N ( I α * u 2 s , α , θ + ε ) u 2 s , α , θ + ε d x λ 2 ( N + α ) 2 q 2 R N ( I α * u q ) u q d x N + α 2 ( 2 s , α * ε ) 2 R N ( I α * u 2 s , α * ε ) u 2 s , α * ε d x λ ( N + α ) ( 2 s , α , θ + ε ) q R N ( I α * u 2 s , α , θ + ε ) u q d x λ ( N + α ) q ( 2 s , α * ε ) R N ( I α * u q ) u 2 s , α * ε d x N + α ( 2 s , α , θ + ε ) ( 2 s , α * ε ) R N ( I α * u 2 s , α , θ + ε ) u 2 s , α * ε d x .

We define

P ε { u W rad s , 2 ( R N , θ ) \ { 0 } : P ε ( u ) = 0 }

and

m ε inf u P ε J ε ( u ) .

The following two inequalities are direct conclusions of Proposition 1.1 and the Sobolev embedding theorem:

(1.7) S s R N u 2 s * d x 2 2 s * u D s , 2 ( R N ) 2 , u D s , 2 ( R N )

and

(1.8) S s , θ R N u 2 s , θ * d x 2 2 s , θ * u W rad s , 2 ( R N , θ ) 2 , u W rad s , 2 ( R N , θ ) ,

where S s and S s , θ are the embedding constants with

S s inf u D s , 2 ( R N ) \ { 0 } u D s , 2 ( R N ) 2 R N u 2 s * d x 2 2 s * .

In terms of the properties of m ε (Lemmas 3.3 and 3.5), we deduce that equation ( C ε ) possesses a nonnegative ground state solution. Passing to the limit as ε 0 + , it can be shown that the ground state solution obtained in equation ( C ε ) is in fact ground state solution of the original equation ( C ) by means of the modified version of the generalized Lions-type theorem (Proposition 3.1).

The remainder of this article is as follows. In Section 2, we study the regularity of weak solution of equation (1.1) and establish the Pohožaev identity. In Section 3, we devote to investigate the existence of ground state solution of equation ( C ).

2 Regularity of solutions and the Pohožaev identity

In this section, we study the regularity of solutions and Pohožaev identity of equation (1.1).

2.1 The nonlocal Brézis-Kato’s type regularity estimate with singular potential

Lemma 2.1

[6] Let N 2 and s 1 2 , 1 . Then there holds

sup x > 0 u ( x ) C x N 2 s 2 u D s , 2 ( R N ) , u D s , 2 ( R N ) .

Lemma 2.2

Let N 2 , s 1 2 , 1 , and θ ( 0 , 2 s ) . We have

R N u 2 p ˜ 2 s , θ * x θ d x C R 2 N u ( x ) u ( y ) 2 x y N + 2 s d x d y ,

where p ˜ = 2 s , θ * ( N θ ) N 2 s .

Proof

Combining Lemma 2.1 and the Hardy inequality [37], we can derive

R N u 2 p ˜ 2 s , θ * x θ d x = R N u 2 p ˜ 2 s , θ * 2 u 2 x θ d x C R N u 2 x θ x 2 p ˜ 2 s , θ * 2 N 2 s 2 d x = C R N u 2 x θ x 2 s θ d x C R 2 N u ( x ) u ( y ) 2 x y N + 2 s d x d y .

The proof is completed.□

Lemma 2.3

[33] Let a , b , c , d [ 1 , ) and ξ [ 0 , 2 ] satisfy

ξ a + 2 ξ b + 1 c + 1 d = 1 + α N .

Assume that β ( 0 , 2 ) satisfying

min ( a , b ) α N 1 c < β < max ( a , b ) 1 1 c , min ( a , b ) α N 1 d < 2 β < max ( a , b ) 1 1 d .

Then, for any G L c ( R N ) , H L d ( R N ) , and u L a ( R N ) L b ( R N ) , it follows that

R N [ I α * ( G u β ) ] ( H u 2 β ) d x C R N G c d x 1 c R N H d d x 1 d R N u a d x ξ a R N u b d x 2 ξ b .

Lemma 2.4

Let N 3 , s 1 2 , 1 , α ( 0 , N ) , θ ( 0 , 2 s ) , and β ( 0 , 2 ) . Assume that G , H L 2 N 2 s , θ * ( N + α ) 2 s , θ * 2 N ( R N ) + L 2 N α + 2 s ( R N ) and ( N + α ) 2 s , θ * 2 N 2 N < β < 2 . Then, for any ε > 0 , there exists C ε , μ > 0 such that

R N [ I α * ( G u β ) ] ( H u 2 β ) d x ε 2 R 2 N u ( x ) u ( y ) 2 x y N + 2 s d x d y + C ε , μ R N u 2 x θ d x , u W rad s , 2 ( R N , θ ) .

Proof

Let G = G ¯ + G ̲ and H = H ¯ + H ̲ , where we have G ¯ , H ¯ L 2 N 2 s , θ * ( N + α ) 2 s , θ * 2 N ( R N ) and G ̲ , H ̲ L 2 N α + 2 s ( R N ) . By applying Lemma 2.3 and taking the parameters a , b , c , d , ξ with different values, we can deduce several inequalities as follows.

(i) Let a = b = 2 N N 2 s , c = d = 2 N α + 2 s , ξ = 0 . We obtain

R N [ I α * ( G ̲ u β ) ] ( H ̲ u 2 β ) d x C R N G ̲ 2 N α + 2 s d x α + 2 s 2 N R N H ̲ 2 N α + 2 s d x α + 2 s 2 N R N u 2 N N 2 s d x N 2 s N .

(ii) Let a = b = 2 s , θ * , c = d = 2 N 2 s , θ * ( N + α ) 2 s , θ * 2 N , and ξ = 2 . One can deduce

R N [ I α * ( G ¯ u β ) ] ( H ¯ u 2 β ) d x C R N G ¯ 2 N 2 s , θ * ( N + α ) 2 s , θ * 2 N d x ( N + α ) 2 s , θ * 2 N 2 N 2 s , θ * R N H ¯ 2 N 2 s , θ * ( N + α ) 2 s , θ * 2 N d x ( N + α ) 2 s , θ * 2 N 2 N 2 s , θ * R N u 2 s , θ * d x 2 2 s , θ * .

(iii) Let a = 2 s , θ * , b = 2 N N 2 s , c = 2 N α + 2 s , d = 2 N 2 s , θ * ( N + α ) 2 s , θ * 2 N , and ξ = 1 . It follows that

R N [ I α * ( G ̲ u β ) ] ( H ¯ u 2 β ) d x C R N G ̲ 2 N α + 2 s d x α + 2 s 2 N R N H ¯ 2 N 2 s , θ * ( N + α ) 2 s , θ * 2 N d x ( N + α ) 2 s , θ * 2 N 2 N 2 s , θ * × R N u 2 s , θ * d x 1 2 s , θ * R N u 2 N N 2 s d x N 2 s 2 N .

(iv) Let a = 2 s , θ * , b = 2 N N 2 s , c = 2 N 2 s , θ * ( N + α ) 2 s , θ * 2 N , d = 2 N α + 2 s , and ξ = 1 . It is easy to see that

R N [ I α * ( G ¯ u β ) ] ( H ̲ u 2 β ) d x C R N G ¯ 2 N 2 s , θ * ( N + α ) 2 s , θ * 2 N d x ( N + α ) 2 s , θ * 2 N 2 N 2 s , θ * R N H ̲ 2 N α + 2 s d x α + 2 s 2 N × R N u 2 s , θ * d x 1 2 s , θ * R N u 2 N N 2 s d x N 2 s 2 N .

Combining (i)(iv), we can show

R N [ I α * ( G u β ) ] ( H u 2 β ) d x C R N G ̲ 2 N α + 2 s d x R N H ̲ 2 N α + 2 s d x α + 2 s 2 N R 2 N u ( x ) u ( y ) 2 x y N + 2 s d x d y + R N G ¯ 2 N 2 s , θ * ( N + α ) 2 s , θ * 2 N d x R N H ¯ 2 N 2 s , θ * ( N + α ) 2 s , θ * 2 N d x ( N + α ) 2 s , θ * 2 N 2 N 2 s , θ * R N u 2 s , θ * d x C R N G ̲ 2 N α + 2 s d x R N H ̲ 2 N α + 2 s d x α + 2 s 2 N R 2 N u ( x ) u ( y ) 2 x y N + 2 s d x d y + R N G ¯ 2 N 2 s , θ * ( N + α ) 2 s , θ * 2 N d x R N H ¯ 2 N 2 s , θ * ( N + α ) 2 s , θ * 2 N d x ( N + α ) 2 s , θ * 2 N 2 N 2 s , θ * R N u 2 x θ d x , u W rad s , 2 ( R N , θ ) .

Since G ̲ , H ̲ L 2 N α + 2 s ( R N ) , we can choose G ̲ , H ̲ satisfying

C R N G ̲ 2 N α + 2 s d x R N H ̲ 2 N α + 2 s d x α + 2 s 2 N ε 2 .

The proof is completed.□

Remark 2.1

If θ = 0 , Lemma 2.4 remains valid. Moreover, for θ > 0 , Lemma 2.4 covers the result of Moroz-Van Schaftingen [33, Lemma 3.2].

Lemma 2.5

Let G , H L 2 N 2 s , θ * ( N + α ) 2 s , θ * 2 N ( R N ) + L 2 N α + 2 s ( R N ) and u W rad s , 2 ( R N , θ ) be a solution of

(2.1) ( Δ ) s u + u x θ = ( I α * G u ) H .

Then, u L p ( R N ) for any p 2 s , θ * , 2 s * ( N θ ) N 2 s .

Proof

By applying Lemma 2.4 with β = 1 , we know that there exists ζ > 0 satisfying

(2.2) R N ( I α * G φ ) H φ d x 1 2 R 2 N φ ( x ) φ ( y ) 2 x y N + 2 s d x d y + ζ 2 R N φ 2 x θ d x , φ W rad s , 2 ( R N , θ ) .

We choose sequences { G n } , { H n } L 2 N α + 2 s ( R N ) , which satisfy

G n G , H n H and G n G , H n H a.e. in R N .

Define a sequence of continuous coercive bilinear functional Ψ n : W rad s , 2 ( R N , θ ) × W rad s , 2 ( R N , θ ) by

Ψ n ( u , v ) R 2 N ( u ( x ) u ( y ) ) ( v ( x ) v ( y ) ) x y N + 2 s d x d y + ζ R N u v x θ d x R N ( I α * G n u ) H n v d x .

From the Lax-Milgram theorem, it is easy to show there exists a unique solution u n of

(2.3) ( Δ ) s u n + ζ u n x θ = [ I α * ( G n u n ) ] H n + ( ζ 1 ) u x θ ,

where u is the given solution of equation (2.1).

We next claim that u n u in W rad s , 2 ( R N , θ ) as n . Multiplying both sides of (2.3) by u n and integrating it over R N , then

u n D s , 2 ( R N ) 2 + ζ u n L 2 ( R N , θ ) 2 = R N [ I α * ( G n u n ) ] ( H n u n ) d x + ( ζ 1 ) R N u n u x θ d x .

By this and recalling (2.2) and the Hölder and Young inequalities, we find that

1 2 u n D s , 2 ( R N ) 2 + ζ 2 u n L 2 ( R N , θ ) 2 ( ζ 1 ) R N u n 2 x θ d x 1 2 R N u 2 x θ d x 1 2 ζ 1 2 ( u n L 2 ( R N , θ ) 2 + u L 2 ( R N , θ ) 2 ) ,

Thus, we obtain

(2.4) u n D s , 2 ( R N ) 2 + u n L 2 ( R N , θ ) 2 C u L 2 ( R N , θ ) 2 ,

which implies that { u n } is bounded in W rad s , 2 ( R N , θ ) . As a result, there exists u ˆ W rad s , 2 ( R N , θ ) such that u n u ˆ in W rad s , 2 ( R N , θ ) and u n u ˆ a.e. in R N . By G n L 2 N α + 2 s ( R N ) , it is easy to verify G n u n is bounded in L 2 N N + α ( R N ) . Then, we have G n u n G u ˆ in L 2 N N + α ( R N ) . Moreover, by H n H and the Lebesgue dominated convergence theorem, one has H n φ H φ in L 2 N N α ( R N ) . It follows that

R N [ I α * ( G n u n ) ] ( H n φ ) d x R N [ I α * ( G u ˆ ) ] ( H φ ) d x , φ C 0 ( R N ) .

Hence, u ˆ is a weak solution of

( Δ ) s u ˆ + ζ u ˆ x θ = [ I α * ( G u ˆ ) ] H + ( ζ 1 ) u x θ .

Since equation (2.3) admits a unique solution, which immediately yields that u = u ˆ .

For κ > 0 , we define u n , κ : R N R by

u n , κ ( x ) = κ , u n κ , u n , κ < u n < κ , κ , u n κ .

We first claim that u n , κ 2 p 2 s , θ * 2 u n , κ W rad s , 2 ( R N , θ ) . For convenience of expression, we set Ω 1 { x R N : u n σ } and Ω 2 { x R N : u n < σ } . With obvious algebraic manipulation, we can write

R 2 N u n , κ ( x ) 2 p 2 s , θ * 2 u n , κ ( x ) u n , κ ( y ) 2 p 2 s , θ * 2 u n , κ ( y ) 2 x y N + 2 s d x d y = 2 Ω 1 Ω 2 u n , κ ( x ) 2 p 2 s , θ * 2 u n , κ ( x ) u n , κ ( y ) 2 p 2 s , θ * 2 u n , κ ( y ) 2 x y N + 2 s d x d y + ( Ω 2 ) 2 u n , κ ( x ) 2 p 2 s , θ * 2 u n , κ ( x ) u n , κ ( y ) 2 p 2 s , θ * 2 u n , κ ( y ) 2 x y N + 2 s d x d y C Ω 1 Ω 2 u n ( x ) u n ( y ) 2 x y N + 2 s d x d y + Ω 2 Ω 2 u n ( x ) u n ( y ) 2 x y N + 2 s d x d y C R 2 N u n ( x ) u n ( y ) 2 x y N + 2 s d x d y

and

R N u n , κ 2 p 2 s , θ * 2 u n , κ 2 x θ d x = Ω 1 u n , κ 2 p 2 s , θ * 2 u n , κ 2 x θ d x + Ω 2 u n , κ 2 p 2 s , θ * 2 u n , κ 2 x θ d x 2 σ 2 2 p 2 s , θ * 2 R N u n 2 x θ d x ,

which implies that u n , κ 2 p 2 s , θ * 2 u n , κ W rad s , 2 ( R N , θ ) .

Let η ( u n , κ ) u n , κ 2 p 2 s , θ * 2 u n , κ . It is easy to see that η is an increasing function, then

0 ( a b ) ( η ( a ) η ( b ) ) , a , b R .

Consider the functions

Λ ( t ) t 2 2 and Γ ( t ) 0 t ( η ( ξ ) ) 1 2 d ξ .

For fixed a , b R satisfying a > b , from the Jensen inequality, we can find that

(2.5) Λ ( a b ) ( η ( a ) η ( b ) ) = ( a b ) ( η ( a ) η ( b ) ) = ( a b ) b a η ( t ) d t = ( a b ) b a ( Γ ( t ) ) 2 d t b a Γ ( t ) d t 2 .

Analogously, we can show that the aforementioned expressions are true for any a < b , which means

(2.6) Λ ( a b ) ( η ( a ) η ( b ) ) Γ ( a ) Γ ( b ) 2 , a , b R .

By using (2.6), there holds

(2.7) Γ ( u n , κ ) ( x ) Γ ( u n , κ ) ( y ) 2 ( u n , κ ( x ) u n , κ ( y ) ) u n , κ ( x ) 2 p 2 s , θ * 2 u n , κ ( x ) u n , κ ( y ) 2 p 2 s , θ * 2 u n , κ ( y ) .

Taking u n , κ 2 p 2 s , θ * 2 u n , κ as a test function of (2.3), we have

(2.8) R 2 N Γ ( u n , κ ) ( x ) Γ ( u n , κ ) ( y ) 2 x y N + 2 s d x d y + ζ R N u n , κ p 2 s , θ * 2 x θ d x R 2 N ( u ( x ) u ( y ) ) ( ( u n , κ 2 p 2 s , θ * 2 u n , κ ) ( x ) ( u n , κ 2 p 2 s , θ * 2 u n , κ ) ( y ) ) x y N + 2 s d x d y + ζ R N u n , κ 2 p 2 s , θ * 2 u n , κ u n x θ d x = R N [ I α * ( G n u n ) ] H n u n , κ 2 p 2 s , θ * 2 u n , κ d x + ( ζ 1 ) R N u n , κ 2 p 2 s , θ * 2 u n , κ u x θ d x .

If p < 2 N 2 s , θ * ( N + α ) 2 s , θ * 2 N , we choose β = 2 s , θ * p in Lemma 2.4, then there exists C > 0 such that

(2.9) R N ( I α * G n u n , κ ) H n u n , κ 2 p 2 s , θ * 2 u n , κ d x R N [ I α * ( G u n , κ ) ] H u n , κ 2 p 2 s , θ * 1 d x 1 2 R 2 N u n , κ ( x ) p 2 s , θ * u n , κ ( y ) p 2 s , θ * 2 x y N + 2 s d x d y + C R N u n , κ p 2 s , θ * 2 x θ d x = 1 2 R 2 N Γ ( u n , κ ) ( x ) Γ ( u n , κ ) ( y ) 2 x y N + 2 s d x d y + C R N u n , κ p 2 s , θ * 2 x θ d x .

Putting together (2.8) with (2.9), we obtain

(2.10) 1 2 R 2 N Γ ( u n , κ ) ( x ) Γ ( u n , κ ) ( y ) 2 x y N + 2 s d x d y C R N u n 2 p 2 s , θ * x θ + u 2 p 2 s , θ * x θ d x + D n , κ I α * ( H n u n 2 p 2 s , θ * 1 ) G n u n d x ,

where

E n , κ { x R N : u n > σ } .

By p < 2 N 2 s , θ * ( N + α ) 2 s , θ * 2 N and Proposition 1.3, one sees that

D n , κ I α * ( H n u n 2 p 2 s , θ * 1 ) G n u n d x C R N H n u n 2 p 2 s , θ * 1 r d x 1 r E n , κ G n u n t d x 1 t ,

where we have 1 r = N + α 2 N + 1 2 s , θ * 1 p and 1 t = N + α 2 N 1 2 s , θ * + 1 p .

By using the Hölder inequality and u n L p ( R N ) , one obtains H n u n 2 p 2 s , θ * 1 L r ( R N ) and G n u n L t ( R N ) . From the Lebesgue dominated convergence theorem, it is not difficult to verify that

lim σ E n , κ I α * ( H n u n 2 p 2 s , θ * 1 ) G n u n d x = 0 .

Inserting the latter relation into (2.10) and combining the Sobolev embedding theorem, we have

(2.11) limsup n R N u n 2 p N 2 s , θ * ( N 2 s ) d x N 2 s N C limsup n R N u n 2 p 2 s , θ * x θ d x .

By N 4 s , α < N ( N 2 s ) 2 θ N ( N θ ) ( N 2 s + θ ) ( N θ ) , and θ ( 0 , 2 s ) , we can easy verify

2 s , θ * < p ˜ < 2 N 2 s , θ * ( N + α ) 2 s , θ * 2 N ,

where p ˜ = 2 s , θ * ( N θ ) N 2 s . Let p = p ˜ in (2.11). Then, by using (2.2), Lemma 2.2 and Fatou’s lemma, we deduce

limsup n R N u n 2 p ˜ N 2 s , θ * ( N 2 s ) d x N 2 s N C limsup n R N u n 2 p ˜ 2 s , θ * x θ d x C limsup n R 2 N u n ( x ) u n ( y ) 2 x y N + 2 s d x d y .

So, we obtain u L p ( R N ) , where p 2 s , θ * , 2 s * ( N θ ) N 2 s . The proof is completed.□

2.2 Regularity of weak solutions and the Pohožaev identity

Lemma 2.6

Let N max { 4 s , 3 } . If

N > 4 s , f o r θ ( 0 , θ ˜ ) , N > N * , f o r θ θ ˜ , 2 s 3 ,

then it follows that

(2.12) N ( N 2 s ) N + 2 ( s θ ) < N ( N 2 s ) 2 θ N ( N θ ) ( N 2 s + θ ) ( N θ ) ,

where N * = 2 ( θ s ) ( θ + 2 s ) + 2 θ ( 2 s θ ) 3 2 s 3 θ and θ ˜ = 8 82 6 159 3 82 + 6 159 3 3 s .

Proof

By a straightforward computation, one easily sees that the inequality (2.12) is equal to

(2.13) ( 2 s 3 θ ) N 2 + 4 ( θ 2 + s θ 2 s 2 ) N + 2 ( 4 s 3 2 s 2 θ θ 3 ) > 0 .

Then, we derive

Δ = 4 ( θ 2 + s θ 2 s 2 ) 2 2 ( 2 s 3 θ ) ( 4 s 3 2 s 2 θ θ 3 ) = 2 θ ( 2 s θ ) 3 > 0 .

We set

N * = 2 ( θ s ) ( θ + 2 s ) + 2 θ ( 2 s θ ) 3 2 s 3 θ N * = 2 ( θ s ) ( θ + 2 s ) 2 θ ( 2 s θ ) 3 2 s 3 θ .

Case 1. If 2 s 3 θ > 0 θ 0 , 2 s 3 , then (2.13) is equal to

N > N * or N < N * .

We first compare N * with 4 s . If 4 s > N * , then by simplifying the operations, we obtain

(2.14) 4 s 2 10 s θ + 2 θ 2 > 2 θ ( 2 s θ ) 3 .

According to the range of θ , we divide Case 1 into the following two subcases.

  1. For θ ( 5 17 ) s 2 , 2 s 3 , one has 4 s 2 10 s θ + 2 θ 2 < 0 , which shows that N * > 4 s .

  2. For θ 0 , ( 5 17 ) s 2 , one has 4 s 2 10 s θ + 2 θ 2 > 0 . If 4 s > N * , by (2.14), then

    4 s 3 18 s 2 θ + 8 s θ 2 θ 3 > 0 .

Let g ( θ ) = θ 3 + 8 s θ 2 18 s 2 θ + 4 s 3 , then g ( θ ) = 3 θ 2 + 16 s θ 18 s 2 . By g max ( θ ) = g 2 s 3 < 0 for θ 0 , 2 s 3 , we know that g ( ) is a decreasing function on θ 0 , 2 s 3 . Furthermore, from g ( 0 ) > 0 and g ( 5 17 ) s 2 < 0 , we affirm that there exists θ ˜ 0 , ( 5 17 ) s 2 such that g ( θ ˜ ) = 0 and

g ( θ ) > 0 for θ ( 0 , θ ˜ ) , g ( θ ) < 0 for θ θ ˜ , ( 5 17 ) s 2 .

In the following, we will work out θ ˜ . According to

θ ˜ 3 8 s θ ˜ 2 + 18 s 2 θ ˜ 4 s 3 = 0 ,

and we set t = θ ˜ 8 s 3 . The aforementioned formulation can transform into

t 3 10 s 2 3 t + 164 s 3 27 = 0 .

It is not difficult to figure out the aforementioned equation, namely,

t * = 82 6 159 3 + 82 + 6 159 3 3 s ,

from which we infer θ ˜ = 8 82 6 159 3 82 + 6 159 3 3 s .

Therefore, we obtain

4 s > N * , for θ ( 0 , θ ˜ ) , N * > 4 s , for θ θ ˜ , 2 s 3 .

We next compare N * with 4 s . From the aforementioned discussion, we just need consider θ θ ˜ , 2 s 3 . If 4 s > N * , then by simplifying the operations one has

4 s 3 18 s 2 θ + 8 s θ 2 θ 3 < 0 .

As mentioned earlier, we know that g ( ) is a decreasing function on θ 0 , 2 s 3 . Thus, we have

g max ( θ ) = g 2 s 3 < 0 ,

which indicates N * < 4 s for θ 0 , 2 s 3 .

Case 2. If 2 s 3 θ < 0 θ 2 s 3 , 2 s , then (2.13) is equivalent to

N * < N < N *

We first compare N * with 4 s . If N * > 4 s , it follows that

4 s 3 18 s 2 θ + 8 s θ 2 θ 3 > 0 .

By the monotonicity of function g ( ) , it is easy to check that g ( ) exists a unique minimum point on 2 s 3 , 2 s , and g max ( θ ) = max { g 2 s 3 , g ( 2 s ) } . Moreover, by

g 2 s 3 < 0 and g ( 2 s ) < 0 ,

one has N * < 4 s . Hence, we have

N * < N < N * < 4 s for θ 2 s 3 , 2 s .

Combined the above two cases and our assumptions on N , our desired result follows. The proof is completed.□

Lemma 2.7

[25] Let 1 r , g L t 1 ( R N ) , and h L t 2 ( R N ) . Then there exists C > 0 such that

g * h L r ( R N ) C g L t 1 ( R N ) h L t 2 ( R N ) ,

where

1 t 1 + 1 t 2 = 1 + 1 r .

Lemma 2.8

Assume that all the conditions described in Theorem 1.1are satisfied. Let u W rad s , 2 ( R N , θ ) be a nontrivial solution of (1.1). Then we have

I α * F ( u ) L ( R N ) C

Proof

By using Lemma 2.5, one has u L p ( R N ) for p 2 s , θ * , 2 s * ( N θ ) N 2 s . From ( F 1 ) , one easily sees that F ( u ) L p ˆ ( R N ) for p ˆ 2 N N + α , 2 N ( N θ ) ( N + α ) ( N 2 s ) .

We choose ε 0 , α ( N 2 α 2 ) ( N 2 s ) N 2 + ( 2 s 2 θ α ) N + 2 s α , then I α can be decomposed by

I α = I α 1 + I α 2 ,

where I α 1 L N ε N α ( R N ) and I α 2 L N + ε N α ( R N ) . Taking r = in Lemma 2.7, then

(2.15) I α 1 * F ( u ) L ( R N ) C I α 1 L N ε N α ( R N ) F ( u ) L N ε α ε ( R N ) .

Similarly, we obtain

(2.16) I α 2 * F ( u ) L ( R N ) C I α 2 L N + ε N α ( R N ) F ( u ) L N + ε α + ε ( R N ) .

By combining ε 0 , α ( N 2 α 2 ) ( N 2 s ) N 2 + ( 2 s 2 θ α ) N + 2 s α and Lemma 2.6, we have

(2.17) 2 N N + α < N + ε α + ε < N ε α ε < 2 N ( N θ ) ( N + α ) ( N 2 s ) .

Putting together with (2.15)–(2.17), one obtains

I α 1 * F ( u ) L ( R N ) and I α 2 * F ( u ) L ( R N ) .

The proof is completed.□

Lemma 2.9

Assume that all the conditions described in Theorem 1.1are satisfied. Let u W rad s , 2 ( R N , θ ) be a nontrivial solution of equation (1.1). For K > 2 , we set

u K ( x ) = K , u ( x ) > K , u ( x ) , u ( x ) K , K , u ( x ) > K .

For τ > 1 , let u ˜ K = u u K 2 ( τ 1 ) . Then for each r [ 2 s , θ * , 2 s * ] , we have

R N u u K τ 1 r d x 2 r C τ 2 R N u 2 s , α , θ 2 u u K τ 1 2 d x + R N u 2 s , α * 2 u u K τ 1 2 d x .

Proof

We first claim that u ˜ K W rad s , 2 ( R N , θ ) . For convenience, we set Ω ˜ 1 { x R N : u K K } and Ω ˜ 2 { x R N : u K > K } . Then, we have

R 2 N u ˜ K ( x ) u ˜ K ( y ) 2 x y N + 2 s d x d y = ( Ω ˜ 1 ) 2 u ˜ K ( x ) u ˜ K ( y ) 2 x y N + 2 s d x d y + 2 Ω 1 Ω 2 u ˜ K ( x ) u ˜ K ( y ) 2 x y N + 2 s d x d y + ( Ω ˜ 2 ) 2 u ˜ K ( x ) u ˜ K ( y ) 2 x y N + 2 s d x d y C K 4 ( τ 1 ) R 2 N u ( x ) u ( y ) 2 x y N + 2 s d x d y

and

R N u ˜ K 2 x θ d x = Ω 1 u ˜ K 2 x θ d x + Ω 2 u ˜ K 2 x θ d x K 4 ( τ 1 ) Ω 1 u 2 x θ d x + Ω 2 u 2 x θ d x K 4 ( τ 1 ) R N u 2 x θ d x ,

from which one at once deduces that u ˜ K W rad s , 2 ( R N , θ ) . Consider the functions

Λ ˜ ( t ) t 2 2 and Θ ˜ ( t ) 0 t [ u ˜ ( ξ ) ] 1 2 d ξ .

Similar to (2.5)–(2.7), we obtain

(2.18) u ˜ K W rad s , 2 ( R N , θ ) 2 Θ ˜ ( u ) D s , 2 ( R N ) 2 + u ˜ K L 2 ( R N , θ ) 2 C R 2 N ( u ( x ) u ( y ) ) ( u ˜ K ( x ) u ˜ K ( y ) ) x y N + 2 s d x d y + R N u u ˜ K x θ d x .

By J ( u ) , φ = 0 , we choose φ = u ˜ K . It follows from (2.18) and the Sobolev embedding theorem that

R N u u K τ 1 r d x 2 r C R 2 N ( u ( x ) u ( y ) ) ( u ˜ K ( x ) u ˜ K ( y ) ) x y N + 2 s d x d y + R N u u ˜ K x θ d x C τ 2 R N u 2 s , α , θ 2 u u K τ 1 2 d x + R N u 2 s , α * 2 u u K τ 1 2 d x .

The proof is completed.□

Thanks to Lemmas 2.8 and 2.9, we can finally prove Theorem 1.1.

Proof of Theorem 1.1

(i) ( L estimate) We split our following discussion into two cases.

Case 1. 2 s , α * 2 N 4 s + α .

In this case, we let τ = 2 s * 2 .

Step 1. It follows from Lemma 2.5 that u L p ( R N ) for p 2 s , θ * , 2 s * ( N θ ) N 2 s . So we have

2 s , θ * < 2 s , α , θ + 2 ( τ 1 ) < 2 s , α * + 2 ( τ 1 ) < 2 s * ( N θ ) N 2 s ,

which yields

R N u 2 s , α , θ 2 u u K τ 1 2 d x < and R N u 2 s , α * 2 u u K τ 1 2 d x < .

For 0 < R < , we define

D τ R N u 2 s , α , θ 2 u u K τ 1 2 d x = { x : u ( x ) R } u 2 s , α , θ 2 u u K τ 1 2 d x + { x : u ( x ) > R } u 2 s , α , θ 2 u u K τ 1 2 d x = D τ ( R ) + D τ c ( R )

and

D ˜ τ R N u 2 s , α * 2 u u K τ 1 2 d x = { x : u ( x ) R } u 2 s , α * 2 u u K τ 1 2 d x + { x : u ( x ) > R } u 2 s , α * 2 u u K τ 1 2 d x = D ˜ τ ( R ) + D ˜ τ c ( R ) .

It is not difficult to see that

lim R D τ ( R ) = D τ , lim R 0 D τ ( R ) = 0

and

lim R D ˜ τ ( R ) = D ˜ τ , lim R 0 D ˜ τ ( R ) = 0 .

Moreover, we have

D τ = D τ ( R ) or D ˜ τ = D ˜ τ ( R ) , R < .

Hence, we obtain u L ( R N ) . Thus, we just need to handle the case

D τ ( R ) < D τ and D ˜ τ ( R ) < D ˜ τ , R < .

For convenience, we set R = 1 . Then, there exist 0 < C 1 , C 2 < satisfying

(2.19) D τ ( 1 ) = C 1 D τ and D ˜ τ ( 1 ) = C 2 D ˜ τ .

According to 2 s , α , θ < 2 , one sees that

D τ c ( 1 ) = { x : u ( x ) > 1 } u 2 s , α , θ 2 u u K τ 1 2 d x { x : u ( x ) > 1 } u u K τ 1 2 d x .

In view of (2.19), we can infer

(2.20) D τ = D τ ( 1 ) + D τ c ( 1 ) = 1 1 C 1 D τ c ( 1 ) 1 1 C 1 { x : u ( x ) > 1 } u u K τ 1 2 d x .

Analogously, we obtain

(2.21) D ˜ τ = D ˜ τ ( 1 ) + D ˜ τ c ( 1 ) = 1 1 C 2 D ˜ τ c ( 1 ) 1 1 C 2 { x : u ( x ) > 1 } u u K τ 1 2 d x .

Putting together (2.20), (2.21) with Lemma 2.9, then

(2.22) R N u u K τ 1 r d x 2 r C 1 C 1 + C 1 C 2 R N u u K τ 1 2 d x .

Taking K in (2.22). It follows that

(2.23) R N u r τ d x 2 r C 1 C 1 + C 1 C 2 R N u 2 τ d x .

By using Lemma 2.9 again, we can deduce

u L r 2 s * 2 ( R N ) C 1 C 1 + C 1 C 2 1 2 s * u L 2 s * ( R N ) < .

By r [ 2 s , θ * , 2 s * ] , one has u L p 1 ( R N ) for p 1 [ 2 s , θ * , 2 s * ] 2 s , θ * 2 s * 2 , ( 2 s * ) 2 2 .

Let r = 2 s * . Then, we have

(2.24) u L ( 2 s * ) 2 2 ( R N ) C 1 C 1 + C 1 C 2 1 2 s * u L 2 s * ( R N ) < .

Step 2. By Step 1, one obtains u L p 1 ( R N ) for p 1 [ 2 s , θ * , 2 s * ] 2 s , θ * 2 s * 2 , ( 2 s * ) 2 2 . By using the fact of

2 s , θ * 2 s * 2 < 2 s , α , θ + 2 ( τ 2 1 ) < 2 s , α * + 2 ( τ 2 1 ) < ( 2 s * ) 2 2 ,

and the definition of D τ and D ˜ τ , it is easy to verify

D τ 2 < and D ˜ τ 2 < .

Similar to Step 1, we just need to consider the case

D τ 2 ( 1 ) < D τ 2 and D ˜ τ 2 ( 1 ) < D ˜ τ 2 .

Let us observe that

(2.25) D τ ( 1 ) = { x : u ( x ) 1 } u 2 s , α , θ 2 u u K τ 1 2 d x { x : u ( x ) 1 } u 2 s , α , θ 2 u u K τ 1 ( u K τ 2 τ ) 2 d x = D τ 2 ( 1 )

and

(2.26) D τ c ( 1 ) = { x : u ( x ) > 1 } u 2 s , α , θ 2 u u K τ 1 2 d x { x : u ( x ) > 1 } u 2 s , α , θ 2 u u K τ 1 ( u K τ 2 τ ) 2 d x = D τ 2 c ( 1 ) .

Taking into account (2.20), (2.21), (2.25), and (2.26), we deduce

D τ 2 ( 1 ) D τ ( 1 ) = C 1 1 C 1 D τ c ( 1 ) C 1 1 C 1 D τ 2 c ( 1 ) ,

which shows

D τ 2 1 1 C 1 { x : u ( x ) > 1 } u u K τ 2 1 2 d x .

Similarly, we have

D ˜ τ 2 1 1 C 2 { x : u ( x ) > 1 } u u K τ 2 1 2 d x .

Combining (2.24), τ 2 = 2 s * 2 2 > 1 and Lemma 2.9, then

u L r 2 s * 2 2 ( R N ) C 1 C 1 + C 1 C 2 1 2 2 s * 2 2 u L ( 2 s * ) 2 2 ( R N ) < ,

which gives u L p 2 ( R N ) for p 2 [ 2 s , θ * , 2 s * ] 2 s , θ * 2 s * 2 2 , 2 s * 2 s * 2 2 .

Let r = 2 s * . By (2.24), we obtain

(2.27) u L ( 2 s * ) 3 2 2 ( R N ) C 1 C 1 + C 1 C 2 1 2 s * + 1 2 2 s * 2 2 u L 2 s * ( R N ) < .

Step 3. Iterating the above process, one sees that

u L r 2 s * 2 n ( R N ) C 1 C 1 + C 1 C 2 1 2 2 s * 2 n u L ( 2 s * ) n 2 n 1 ( R N ) .

Let r = 2 s * . Repeating the steps in (2.24) and (2.27), then

(2.28) u L ( 2 s * ) n + 1 2 n ( R N ) C 1 C 1 + C 1 C 2 i = 1 n 1 2 2 s * 2 i u L 2 s * ( R N ) .

So, we obtain

lim i 2 2 s * 2 i 2 2 s * 2 i + 1 = 2 2 s * < 1 .

Thus, the series i = 1 n 1 2 2 s * 2 i converges absolutely. Taking n in (2.28), we have

u L ( R N ) C u L 2 s * ( R N ) < .

Case 2. 2 s , α * > 2 N < 4 s + α .

Step 1. Let τ 1 1 + 2 s , θ * 2 s , α , θ 2 , 1 + 2 s * ( N θ ) N 2 s 2 s , α , θ 2 . We first prove

(2.29) 1 + R N u 2 s , α * τ 1 d x 2 2 s , α * ( τ 1 1 ) < .

By the definition of u K , we deduce

R N u 2 s , α , θ 2 u u K τ 1 1 2 d x R N u 2 τ 1 + 2 s , α , θ 2 d x .

Let d > 0 . Using the Hölder inequality, one has

(2.30) R N u 2 s , α * 2 u u K τ 1 1 2 d x d 2 s , α * 2 s , α , θ { x : u ( x ) d } u 2 s , α , θ 2 u u K τ 1 1 2 d x + { x : u ( x ) > d } u 2 s , α * 2 u u K τ 1 1 2 d x d 2 s , α * 2 s , α , θ R N u 2 s , α , θ + 2 ( τ 1 1 ) d x + { x : u ( x ) > d } u 2 s , α * d x 2 s , α * 2 2 s , α * R N u u K τ 1 1 2 s , α * d x 2 2 s , α * .

By u L p ( R N ) for p 2 s , θ * , 2 s * ( N θ ) N 2 s , then we can fix d such that

(2.31) { x : u ( x ) > d } u 2 s , α * d x 2 s , α * 2 2 s , α * 1 2 C τ 1 2 .

Combined (2.30)–(2.31) with Lemma 2.9, then

R N u u K τ 1 1 2 s , α * d x 2 2 s , α * 2 C τ 1 2 R N u 2 s , α , θ 2 u u K τ 1 1 2 d x + d 2 s , α * 2 s , α , θ R N u 2 s , α , θ + 2 ( τ 1 1 ) d x .

Let K . The latter formula turns into

(2.32) R N u 2 s , α * τ 1 d x 2 2 s , α * 2 C τ 1 2 ( 1 + l 2 s , α * 2 s , α , θ ) R N u 2 s , α , θ + 2 ( τ 1 1 ) d x .

Since 2 s , α , θ + 2 ( τ 1 1 ) 2 s , θ * , 2 s * ( N θ ) N 2 s , by (2.32), we deduce (2.29). So u L p ˆ 1 ( R N ) , where

p ˆ 1 2 s , θ * , 2 s * ( N θ ) N 2 s 2 s , α * 1 + 2 s , θ * 2 s , α , θ 2 , 2 s , α * 1 + 2 s * ( N θ ) N 2 s 2 s , α , θ 2 .

Step 2. Let τ 2 = 1 + 2 s , α * 2 ( τ 1 1 ) . We next prove

1 + R N u 2 s , α * τ 2 d x 2 2 s , α * ( τ 2 1 ) ( C τ 2 ) 2 τ 2 1 1 + R N u 2 s , α * τ 1 d x 2 2 s , α * ( τ 1 1 ) .

Let τ [ τ 1 , τ 2 ] . Then

2 s , θ * 2 s , α , θ + 2 ( τ 1 ) < 2 s , α * + 2 ( τ 1 ) 2 s , α * τ 1 .

Combining Lemma 2.9 and (2.29), we observe

(2.33) R N u 2 s , α * τ d x 2 2 s , α * C τ 2 R N u 2 s , α , θ + 2 ( τ 1 ) d x + R N u 2 s , α * + 2 ( τ 1 ) d x < .

Let τ = τ 2 in (2.33). It follows that

R N u 2 s , α * τ 2 d x 2 2 s , α * C τ 2 2 R N u 2 s , α , θ + 2 ( τ 2 1 ) d x + R N u 2 s , α * + 2 ( τ 2 1 ) d x < .

In view of Young’s inequality, we obtain

R N u 2 s , α , θ + 2 ( τ 2 1 ) d x = R N u a 2 u b 2 d x a 2 2 s , α * R N u 2 s , α * d x + 2 s , α * a 2 2 s , α * R N u 2 s , α * + 2 ( τ 2 1 ) d x C 1 + R N u 2 s , α * + 2 ( τ 2 1 ) d x ,

where a 2 = 2 s , α * ( 2 s , α * 2 s , α , θ ) 2 ( τ 2 1 ) and b 2 = 2 s , α , θ + 2 ( τ 2 1 ) 2 s , α * ( 2 s , α * 2 s , α , θ ) 2 ( τ 2 1 ) . So

R N u 2 s , α * τ 2 d x 2 2 s , α * C τ 2 2 1 + R N u 2 s , α * + 2 ( τ 2 1 ) d x .

Owing to 2 s , α * > 2 , one sees that

( x 1 + x 2 ) 2 2 s , α * x 1 2 2 s , α * + x 2 2 2 s , α * , x 1 , x 2 > 0 .

Then, there holds

1 + R N u 2 s , α * τ 2 d x 2 2 s , α * 1 + R N u 2 s , α * τ 2 d x 2 2 s , α * C τ 2 2 1 + R N u 2 s , α * + 2 ( τ 2 1 ) d x ,

from which we can deduce

1 + R N u 2 s , α * τ 2 d x 2 2 s , α * ( τ 2 1 ) ( C τ 2 ) 2 τ 2 1 1 + R N u 2 s , α * τ 1 d x 1 τ 2 1 = ( C τ 2 ) 2 τ 2 1 1 + R N u 2 s , α * τ 1 d x 2 2 s , α * ( τ 1 1 ) .

Step 3. We iterate the above process and set

(2.34) τ i + 1 1 = 2 s , α * 2 ( τ i 1 ) , i 1 .

Thus, we have

1 + R N u 2 s , α * τ i + 1 d x 2 2 s , α * ( τ i + 1 1 ) ( C τ i + 1 ) 2 τ i + 1 1 1 + R N u 2 s , α * τ i d x 1 τ i + 1 1 = ( C τ i + 1 ) 2 τ i + 1 1 1 + R N u 2 s , α * τ i d x 2 2 s , α * ( τ i 1 ) .

Then, it follows that

R N u 2 s , α * τ n + 1 d x 2 2 s , α * ( τ n + 1 1 ) 1 + R N u 2 s , α * τ n + 1 d x 2 2 s , α * ( τ n + 1 1 ) i = 1 n ( C τ i + 1 ) 2 τ i + 1 1 1 + R N u 2 s , α * τ 1 d x 2 2 s , α * ( τ 1 1 ) ,

which implies

(2.35) u L 2 s , α * τ n + 1 ( R N ) i = 1 n ( C τ i + 1 ) 2 τ i + 1 1 1 + R N u 2 s , α * τ 1 d x 2 2 s , α * ( τ 1 1 ) τ n + 1 1 2 τ n + 1 .

By using (2.34), then

(2.36) τ n + 1 = 1 + 2 s , α * 2 n ( τ 1 1 ) .

From (2.35) and (2.36), we deduce

(2.37) u L 2 s , α * τ n + 1 ( R N ) i = 1 n ( C τ i + 1 ) 2 τ i + 1 1 1 + R N u 2 s , α * τ 1 d x 2 2 s , α * ( τ 1 1 ) ( 2 s , α * ) n ( τ 1 1 ) 2 [ 2 n + ( 2 s , α * ) n ( τ 1 1 ) ] .

It is not difficult to see that

(2.38) lim n i = 1 n ( C τ i + 1 ) 2 τ i + 1 1 = lim n e 2 i = 1 n ln C τ i + 1 1 + ln τ i + 1 τ i + 1 1 .

Since

(2.39) lim i ln C τ i + 1 1 i = lim i 2 i ln C ( 2 s , α * ) i ( τ 1 1 ) i = 2 2 s , α * < 1 ,

the series i = 1 ln C τ i + 1 1 converges absolutely.

Moreover, it follows that

(2.40) lim i ln τ i + 2 τ i + 2 1 τ i + 1 1 ln τ i + 1 = 2 2 s , α * lim i ln 1 + 2 s , α * 2 ( τ i + 1 1 ) ln τ i + 1 2 2 s , α * lim i ln 2 s , α * 2 + 2 s , α * 2 ( τ i + 1 1 ) ln τ i + 1 = 2 2 s , α * lim i ln 2 s , α * 2 ln τ i + 1 + ln τ i + 1 ln τ i + 1 < 1 .

Then, we know that the series i = 1 ln τ i + 1 τ i + 1 1 converges absolutely.

According to (2.38)–(2.40), we have i = 1 ( C τ i + 1 ) 2 τ i + 1 1 < . Taking n in (2.35), one has

u L ( R N ) < .

(ii) (Pohožaev identity) Combining Theorem 1.1 (i), Lemmas 2.8,2.9 and ( F 1 ) , there exists C ( Ω ) > 0 such that

( Δ ) s u = h ( u ) C ( Ω ) ( u 2 s , α , θ 2 u + u 2 s , α * 2 u ) , Ω R N \ { 0 } ,

where h ( u ) u x θ + ( I α * F ( u ) ) f ( u ) . Obviously, one has h L ( Ω ) . From [44, Theorem1.2], there exists σ ( 0 , 1 s ) and C > 0 , depending on Ω and s such that u satisfies u δ ( x ) s C 0 , σ ( Ω ¯ ) and

u δ ( x ) s C 0 , σ ( Ω ¯ ) C h L ( Ω ) ,

where δ ( x ) s = dist ( x , Ω ) s . Let ω satisfy Δ ω = h ( u ) C 0 , σ ( Ω ¯ ) . By the Hölder regularity theory for Laplacian, we obtain ω C 2 , σ ( Ω ¯ ) . It follows from 2 s + σ > 1 that ( Δ ) 1 s ω C 1 , 2 s + σ 1 ( Ω ¯ ) . By ( Δ ) s ( u ( Δ ) 1 s ω ) = 0 , we know that u has the same regularity as ( Δ ) 1 s ω . Thus, we deduce u C 1 , 2 α + σ 1 ( Ω ¯ ) . Since the corresponding Hölder norms depend only on N and s , we can derive these estimates are global in R N \ { 0 } .

Similar to [33, Theorem 3], we can prove the following Pohožaev identity

N 2 s 2 u D s , 2 ( R N ) 2 + N θ 2 u L 2 ( R N , θ ) 2 = N + α 2 R N ( I α * F ( u ) ) F ( u ) d x .

This ends the proof.□

3 Proof of Theorem 1.2

In this section, we will prove Theorem 1.2 by virtue of subcritical approximation arguments, the Pohožaev manifold method and the modified version of the generalized Lions-type theorem.

As a consequence of Theorem 1.1, we introduce the following Pohožaev manifold

P { u W rad s , 2 ( R N , θ ) \ { 0 } : P ( u ) = 0 } ,

where

P ( u ) = N 2 s 2 u D s , 2 ( R N ) 2 + N θ 2 u L 2 ( R N , θ ) 2 N + α 2 ( 2 s , α , θ ) 2 R N ( I α * u 2 s , α , θ ) u 2 s , α , θ d x λ 2 ( N + α ) 2 q 2 R N ( I α * u q ) u q d x N + α 2 ( 2 s , α * ) 2 R N ( I α * u 2 s , α * ) u 2 s , α * d x λ ( N + α ) 2 s , α , θ q R N ( I α * u 2 s , α , θ ) u q d x λ ( N + α ) q 2 s , α * R N ( I α * u q ) u 2 s , α * d x N + α 2 s , α , θ 2 s , α * R N ( I α * u 2 s , α , θ ) u 2 s , α * d x .

For each u P , we define

m inf u P J ( u ) .

Lemma 3.1

Assume that C 1 , C 2 , C 3 > 0 , s 1 2 , 1 , α ( 0 , N ) , and θ ( 0 , 2 s ) . Define k : R + R as follows:

k ( t ) = C 1 t N 2 s + C 2 t N θ C 3 t N + α .

Then k ( t ) has a unique critical point, which corresponds to its maximum.

Proof

Obviously, we have

k ( t ) = C 1 ( N 2 s ) t N 2 s 1 + C 2 ( N θ ) t N θ 1 C 3 ( N + α ) t N + α 1 .

Since one can easily see that k ( t ) > 0 for t > 0 small and k ( t ) < 0 as t . This means that k ( t ) has at least one maximum point. We claim that the maximum point is unique. On the contrary, suppose that there exists two different critical points t 1 , t 2 > 0 such that

k ( t 1 ) = C 1 ( N 2 s ) t 1 N 2 s 1 + C 2 ( N θ ) t 1 N θ 1 C 3 ( N + α ) t 1 N + α 1 = 0

and

k ( t 2 ) = C 1 ( N 2 s ) t 2 N 2 s 1 + C 2 ( N θ ) t 2 N θ 1 C 3 ( N + α ) t 2 N + α 1 = 0 .

From this, we conclude that

C 1 ( N 2 s ) ( t 1 θ 2 s t 2 θ 2 s ) = C 3 ( N + α ) ( t 1 α + θ t 2 α + θ ) .

This implies that t 1 = t 2 . The proof is completed.□

Lemma 3.2

Assume that all conditions described in Theorem 1.2are satisfied. Then, there exists a unique t u > 0 such that P ( u t u ) = 0 , where u t u x t . Moreover, J ( u t u ) = max t 0 J ( u t ) .

Proof

We set η ( t ) = J ( u t ) . It is easy to observe that

η ( t ) = t N 2 s 2 u D s , 2 ( R N ) 2 + t N θ 2 u L 2 ( R N , θ ) 2 t N + α 2 ( 2 s , α , θ ) 2 R N ( I α * u 2 s , α , θ ) u 2 s , α , θ d x λ 2 t N + α 2 q 2 R N ( I α * u q ) u q d x t N + α 2 ( 2 s , α * ) 2 R N ( I α * u 2 s , α * ) u 2 s , α * d x λ t N + α 2 s , α , θ q R N ( I α * u 2 s , α , θ ) u q d x λ t N + α q 2 s , α * R N ( I α * u q ) u 2 s , α * d x t N + α 2 s , α , θ 2 s , α * R N ( I α * u 2 s , α , θ ) u 2 s , α * d x

and

(3.1) η ( t ) = ( N 2 s ) t N 2 s 1 2 u D s , 2 ( R N ) 2 + ( N θ ) t N θ 1 2 u L 2 ( R N , θ ) 2 ( N + α ) t N + α 1 2 ( 2 s , α , θ ) 2 R N ( I α * u 2 s , α , θ ) u 2 s , α , θ d x λ 2 ( N + α ) t N + α 1 2 q 2 R N ( I α * u q ) u q d x ( N + α ) t N + α 1 2 ( 2 s , α * ) 2 R N ( I α * u 2 s , α * ) u 2 s , α * d x λ ( N + α ) t N + α 1 2 s , α , θ q R N ( I α * u 2 s , α , θ ) u q d x λ ( N + α ) t N + α 1 q 2 s , α * R N ( I α * u q ) u 2 s , α * d x λ ( N + α ) t N + α 1 2 s , α , θ 2 s , α * R N ( I α * u 2 s , α , θ ) u 2 s , α * d x .

By P ( u t ) = t η ( t ) = 0 , the desired results follow by (3.1) and Lemma 3.1 immediately. The proof is completed.□

Lemma 3.3

Assume that all conditions described in Theorem 1.2are satisfied. Then

0 < m < m * min { m 1 , m 2 } ,

where

m 1 α + 2 s 2 ( N + α ) ( 2 s , α , θ ) 2 ( N 2 s ) C N , α ( N + α ) 2 N 2 s , θ * ( N + α ) 2 N S s , θ 2 s , θ * ( N + α ) 2 s , θ * ( N + α ) 2 N

and

m 2 α + 2 s 2 ( N + α ) ( 2 s , α * ) 2 ( N 2 s ) C N , α ( N + α ) 2 N 2 s * ( N + α ) 2 N S s 2 s * ( N + α ) 2 s * ( N + α ) 2 N .

Proof

To do this, we choose v W rad s , 2 ( R N , θ ) satisfying

v W rad s , 2 ( R N , θ ) 2 = 1 and R N ( I α * v q ) v q d x > 0 .

By using Lemma 3.2, we can find t v , λ > 0 such that

J ( v t v , λ ) = sup t 0 J ( v t ) and P ( v t v , λ ) = 0 .

Consequently, one has

(3.2) N 2 s 2 t v , λ N 2 s v D s , 2 ( R N ) 2 + N θ 2 t v , λ N θ v L 2 ( R N , θ ) 2 = t v , λ N + α N + α 2 ( 2 s , α , θ ) 2 R N ( I α * v 2 s , α , θ ) v 2 s , α , θ d x + λ 2 ( N + α ) 2 q 2 R N ( I α * v q ) v q d x + N + α 2 ( 2 s , α * ) 2 R N ( I α * v 2 s , α * ) v 2 s , α * d x + λ ( N + α ) 2 s , α , θ q R N ( I α * v 2 s , α , θ ) v q d x + λ ( N + α ) q 2 s , α * R N ( I α * v q ) v 2 s , α * d x + N + α 2 s , α , θ 2 s , α * R N ( I α * v 2 s , α , θ ) v 2 s , α * d x .

For this, there holds

N 2 s 2 t v , λ N 2 s v D s , 2 ( R N ) 2 + N θ 2 t v , λ N θ v L 2 ( R N , θ ) 2 N + α 2 ( 2 s , α * ) 2 t v , λ N + α R N ( I α * v 2 s , α * ) v 2 s , α * d x .

Thus, we know that { t v , λ } is bounded.

Next, we verify t v , λ 0 as λ . Argue by contradiction, there exist t ˆ > 0 and a sequence { λ n } with λ n as n , such that t v , λ n t ˆ . Hence, letting n , we have

λ n 2 ( N + α ) 2 q 2 t ˆ N + α R N ( I α * v q ) v q d x .

Letting n in (3.2), we deduce

N θ 2 t ˆ N θ v W rad s , 2 ( R N , θ ) 2 lim n λ n 2 ( N + α ) 2 q 2 t v , λ n 2 q + 1 R N ( I α * v q ) v q d x = .

Obviously, the latter relation yields a contradiction with v is bounded. So t v , λ 0 as λ . Therefore, we can infer

lim λ sup t 0 J ( v t ) = lim λ J ( v t v , λ ) = 0 .

Hence, we observe that there exists 0 < Λ < , such that for any λ > Λ , we arrive at

sup t 0 J ( v t ) < m * .

The proof is completed.□

Lemma 3.4

Assume that all conditions described in Theorem 1.2are satisfied. Then, there exists a unique t u > 0 such that P ε ( u t u ) = 0 , where u t u x t . Moreover, J ε ( u t u ) = max t 0 J ε ( u t ) .

The proof is similar to Lemma 3.2. We omit any further detail.

Lemma 3.5

Assume that all conditions described in Theorem 1.2are satisfied. Then

  1. inf u P ε J ε ( u ) = m ε = inf u W rad s , 2 ( R N , θ ) \ { 0 } max t 0 J ε ( u t ) > 0 .

  2. limsup ε 0 + m ε m .

Proof

(i) By virtue of Lemma 3.4, we obtain

m ε = inf u W rad s , 2 ( R N , θ ) \ { 0 } max t 0 J ε ( u t ) .

We claim that m ε > 0 . Indeed, for u P ε , by Proposition 1.3, one sees that

N 2 s 2 u D s , 2 ( R N ) 2 + N θ 2 u L 2 ( R N , θ ) 2 = N + α 2 ( 2 s , α , θ + ε ) 2 R N ( I α * u 2 s , α , θ + ε ) u 2 s , α , θ + ε d x + λ 2 ( N + α ) 2 q 2 R N ( I α * u q ) u q d x + N + α 2 ( 2 s , α * ε ) 2 R N ( I α * u 2 s , α * ε ) u 2 s , α * ε d x + λ ( N + α ) ( 2 s , α , θ + ε ) q R N ( I α * u 2 s , α , θ + ε ) u q d x + λ ( N + α ) q ( 2 s , α * ε ) R N ( I α * u q ) u 2 s , α * ε d x + N + α ( 2 s , α , θ + ε ) ( 2 s , α * ε ) R N ( I α * u 2 s , α , θ + ε ) u 2 s , α * ε d x C 1 u W rad s , 2 ( R N , θ ) 2 s , θ * ( N + α ) N + 2 ε + C 2 u W rad s , 2 ( R N , θ ) 2 q + C 3 u W rad s , 2 ( R N , θ ) 2 s * ( N + α ) N 2 ε + C 4 u W rad s , 2 ( R N , θ ) 2 s , θ * ( N + α ) + 2 N q 2 N + ε + C 5 u W rad s , 2 ( R N , θ ) 2 s * ( N + α ) + 2 N q 2 N ε + C 6 u W rad s , 2 ( R N , θ ) ( 2 s , θ * + 2 s * ) ( N + α ) 2 N .

From the latter expressions, it is clear that u W rad s , 2 ( R N , θ ) C . Consequently, we have

(3.3) J ε ( u ) = J ε ( u ) 1 N + α P ε ( u ) = α + 2 s 2 ( N + α ) u D s , 2 ( R N ) 2 + α + θ 2 ( N + α ) u L 2 ( R N , θ ) 2 C ,

which leads to m ε > 0 .

(ii) For ε 0 , 1 2 , we can find u W rad s , 2 ( R N , θ ) \ { 0 } such that P ( u ) = 0 and J ( u ) < m + ε . Thus, we know that there exists t ˆ > 0 large satisfying

(3.4) J ( u t ˆ ) = t ˆ N 2 s 2 u D s , 2 ( R N ) 2 + t ˆ N θ 2 u L 2 ( R N , θ ) 2 t ˆ N + α 2 ( 2 s , α , θ ) 2 R N ( I α * u 2 s , α , θ ) u 2 s , α , θ d x λ 2 t ˆ N + α 2 q 2 R N ( I α * u q ) u q d x t ˆ N + α 2 ( 2 s , α * ) 2 R N ( I α * u 2 s , α * ) u 2 s , α * d x λ t ˆ N + α 2 s , α , θ q R N ( I α * u 2 s , α , θ ) u q d x λ t ˆ N + α q 2 s , α * R N ( I α * u q ) u 2 s , α * d x t ˆ N + α 2 s , α , θ 2 s , α * R N ( I α * u 2 s , α , θ ) u 2 s , α * d x 1 .

By using the Young inequality, one can see that

(3.5) u 2 s , α , θ + ε 2 s , α * 2 s , α , θ ε 2 s , α * 2 s , α , θ u 2 s , α , θ + ε 2 s , α * 2 s , α , θ u 2 s , α *

and

(3.6) u 2 s , α * ε ε 2 s , α * 2 s , α , θ u 2 s , α , θ + 2 s , α * 2 s , α , θ ε 2 s , α * 2 s , α , θ u 2 s , α * .

Let us define

G ( t , ε ) t N + α 2 ( 2 s , α , θ + ε ) 2 R N ( I α * u 2 s , α , θ + ε ) u 2 s , α , θ + ε d x + λ 2 t N + α 2 q 2 R N ( I α * u q ) u q d x + t N + α 2 ( 2 s , α * ε ) 2 R N ( I α * u 2 s , α * ε ) u 2 s , α * ε d x + λ t N + α ( 2 s , α , θ + ε ) q R N ( I α * u 2 s , α , θ + ε ) u q d x + λ t N + α q ( 2 s , α * ε ) R N ( I α * u q ) u 2 s , α * ε d x + t N + α ( 2 s , α , θ + ε ) ( 2 s , α * ε ) R N ( I α * u 2 s , α , θ + ε ) u 2 s , α * ε d x .

By (3.5), (3.6) and the Lebesgue dominated convergence theorem, it is not difficult to see G ( , ) is continuous on ( t , ε ) [ 0 , t ˆ ] × 0 , 2 s , α * 2 s , α , θ 2 . Therefore, there exits ε ˆ > 0 such that

(3.7) J ε ( u t ) J ( u t ) = G ( t , ε ) G ( t , 0 ) < ε ,

for ( t , ε ) [ 0 , t ˆ ] × ( 0 , ε ˆ ) . Then, it follows from (3.4) and (3.7) that

J ε ( u t ˆ ) < 1 2 , ε ( 0 , ε ˆ ) .

Since by (3.4), we have J ε ( u t ) > 0 for t > 0 small and J ε ( u t ) < 0 for t > 0 large. Obviously, we can find t ε ( 0 , t ˆ ) satisfying d J ε ( u t ) d t t = t ε = 0 . Putting this together with the fact P ε ( u t ε ) = t d J ε ( u t ) d t t = t ε , one sees that P ε ( u t ε ) = 0 . Moreover, by P ( u ) = 0 and Lemma 3.4, one observes J ( u t ε ) J ( u ) . Therefore,

m ε J ε ( u t ε ) < J ( u t ε ) + ε J ( u ) + ε < m + 2 ε ,

for ε ( 0 , ε ˆ ) . The proof is completed.□

Lemma 3.6

Assume that all the conditions described in Theorem 1.2are satisfied. Let { ε n } 0 and { u ε n } be a sequence that satisfies J ε n ( u ε n ) = m ε n . If u ε n u 0 0 , then u 0 is a nontrivial solution of equation ( C ).

Proof

Clearly, we have

J ε n ( u ε n ) = m ε n and P ε n ( u ε n ) = 0 .

In terms of (3.3), it is standard to show { u ε n } is uniformly bounded in W rad s , 2 ( R N , θ ) . Then, up to subsequences,

u ε n u 0 in W rad s , 2 ( R N , θ ) , u ε n u 0 in L r ( R N ) for r ( 2 s , θ * , 2 s * ) and u ε n u 0 a.e. in R N .

To finish the proof, it suffices to show, for φ W rad s , 2 ( R N , θ ) , that

(3.8) R N ( I α * u ε n 2 s , α , θ + ε n ) u ε n 2 s , α , θ + ε n 2 u ε n φ d x = R N ( I α * u 0 2 s , α , θ ) u 0 2 s , α , θ 2 u 0 φ d x ,

(3.9) R N ( I α * u ε n 2 s , α * ε n ) u ε n 2 s , α * ε n 2 u ε n φ d x = R N ( I α * u 0 2 s , α * ) u 0 2 s , α * 2 u 0 φ d x ,

(3.10) R N ( I α * u ε n 2 s , α , θ + ε n ) u ε n 2 s , α * ε n 2 u ε n φ d x = R N ( I α * u 0 2 s , α , θ ) u 0 2 s , α * 2 u 0 φ d x ,

(3.11) R N ( I α * u ε n 2 s , α , θ + ε n ) u ε n q 2 u ε n φ d x = R N ( I α * u 0 2 s , α , θ ) u 0 q 2 u 0 φ d x ,

and

(3.12) R N ( I α * u ε n 2 s , α * ε n ) u ε n q 2 u ε n φ d x = R N ( I α * u 0 2 s , α * ) u 0 q 2 u 0 φ d x .

Since the proof of (3.8)–(3.12) are similar, we just show (3.8). For each ε > 0 , there exists R > 0 large such that

(3.13) lim n { x : x > R } [ ( I α * u ε n 2 s , α , θ + ε n ) u ε n 2 s , α , θ + ε n 2 u ε n φ ( I α * u 0 2 s , α , θ ) u 0 2 s , α , θ 2 u 0 φ ] d x lim n { x : x > R } ( I α * u ε n 2 s , α , θ + ε n ) u ε n 2 s , α , θ + ε n 1 φ d x + lim n { x : x > R } ( I α * u 0 2 s , α , θ ) u 0 2 s , α , θ 1 φ d x lim n { x : x > R } u ε n 2 s , θ * + 2 N ε n N + α d x N + α 2 N { x : x > R } u ε n 2 s , θ * + 2 N ε n 2 s , θ * 2 s , θ * ( N + α ) 2 N 2 s , θ * ( N + α ) 2 N 2 N 2 s , θ * { x : x > R } φ 2 s , θ * d x 1 2 s , θ * + { x : x > R } u 0 2 s , θ * d x N + α 2 N { x : x > R } u 0 2 s , θ * d x 2 s , θ * ( N + α ) 2 N 2 N 2 s , θ * { x : x > R } φ 2 s , θ * d x 1 2 s , θ * < ε 2 .

Making use of Young’s inequality, we obtain

(3.14) R N u ε n 2 s , θ * + 2 N ε n N + α d x ( N + α ) ( 2 s * 2 s , θ * ) 2 N ε n ( N + α ) ( 2 s * 2 s , θ * ) R N u ε n 2 s , θ * d x + 2 N ε n ( N + α ) ( 2 s * 2 s , θ * ) R N u ε n 2 s * d x .

As a result, { u ε n } is bounded in L 2 s , α , θ + ε n ( R N ) .

In light of u ε n u 0 a.e. in R N , for each ε > 0 , there exists n ˜ > 0 , depending on ε , such that, for all n > n ˜ ,

(3.15) u ε n 2 s , α , θ u 0 2 s , α , θ 2 s , α , θ u ε n u 0 u ε n 2 s , α , θ 1 + u 0 2 s , α , θ 1 < ε 2 .

Recalling u ε n L ( R N ) and u ε n u 0 a.e. in R N , we obtain u 0 L ( R N ) . Thus, one has

(3.16) u ε n 2 s , α * u 0 2 s , α , θ u ε n 2 s , α * u 0 2 s , α , θ L ( R N ) u ε n 2 s , α * L ( R N ) + u 0 2 s , α , θ L ( R N ) C .

For ε n > 0 small, gathering together estimates (3.15) and (3.16), we obtain

u ε n 2 s , α , θ + ε n u 0 2 s , α , θ 2 s , α * 2 s , α , θ ε n 2 s , α * 2 s , α , θ u ε n 2 s , α , θ + ε n 2 s , α * 2 s , α , θ u ε n 2 s , α * u 0 2 s , α , θ 2 s , α * 2 s , α , θ ε n 2 s , α * 2 s , α , θ u ε n 2 s , α , θ u 0 2 s , α , θ + ε n 2 s , α * 2 s , α , θ u ε n 2 s , α * u 0 2 s , α , θ < ε .

From this, we obtain u ε n 2 s , α , θ + ε n u 0 2 s , α , θ a.e. in R N . Owing to pointwise convergence of a bounded sequence implies weak convergence (see [43]), we deduce

(3.17) u ε n 2 s , α , θ + ε n u 0 2 s , α , θ in L 2 N N + α ( R N ) .

It follows from Proposition 1.3 that the Riesz potential defines a linear continuous map from L 2 N N + α ( R N ) to L 2 N N α ( R N ) . By exploiting (3.17), we notice that

(3.18) I α * u ε n 2 s , α , θ + ε n I α * u 0 2 s , α , θ in L 2 N N α ( R N ) .

Analogous to (3.14), we immediately have

Ω u ε n 2 s , θ * + 2 N ε n 2 s , θ * 2 s , θ * ( N + α ) 2 N d x 2 s , θ * ( N + α ) 2 N 2 s , θ * ( N + α ) C .

Then, for each ε > 0 , there exists δ > 0 such that when Ω { x R N : x R } and Ω < δ , we derive

Ω u ε n 2 s , α , θ + ε n 2 u ε n φ 2 N N + α d x Ω u ε n 2 s , θ * + 2 N ε n 2 s , θ * 2 s , θ * ( N + α ) 2 N d x 2 s , θ * ( N + α ) 2 N 2 s , θ * ( N + α ) Ω φ 2 s , θ * d x 2 N 2 s , θ * ( N + α ) ε 2 .

We observe that u ε n 2 s , α , θ + ε n 2 u ε n φ 2 N N + α u 0 2 s , α , θ 2 u 0 φ 2 N N + α a.e. in R N , then we obtain

(3.19) lim n { x : x R } u ε n 2 s , α , θ + ε n 2 u ε n φ 2 N N + α d x = { x : x R } u 0 2 s , α , θ 2 u 0 φ 2 N N + α d x .

Considering (3.18) and (3.19), one can infer

(3.20) lim n { x : x R } ( I α * u ε n 2 s , α , θ + ε n ) u ε n 2 s , α , θ + ε n 2 u ε n φ d x = { x : x R } ( I α * u 0 2 s , α , θ ) u 0 2 s , α , θ 2 u 0 φ d x .

From (3.13) and (3.20), we have that

lim n R N ( I α * u ε n 2 s , α , θ + ε n ) u ε n 2 s , α , θ + ε n 2 u ε n φ d x R N ( I α * u 0 2 s , α , θ ) u 0 2 s , α , θ 2 u 0 φ d x lim n { x : x R } ( I α * u ε n 2 s , α , θ + ε n ) u ε n 2 s , α , θ + ε n 2 u ε n φ ( I α * u 0 2 s , α , θ ) u 0 2 s , α , θ 2 u 0 φ d x + lim n { x : x > R } ( I α * u ε n 2 s , α , θ + ε n ) u ε n 2 s , α , θ + ε n 2 u ε n φ ( I α * u 0 2 s , α , θ ) u 0 2 s , α , θ 2 u 0 φ d x < ε .

Similarly, we can prove (3.9)–(3.12). Hence, for each φ W rad s , 2 ( R N , θ ) , we obtain

0 = lim n J ε n ( u ε n ) , φ = lim n R 2 N ( u ε n ( x ) u ε n ( y ) ) ( φ ( x ) φ ( y ) ) x y N + 2 s d x d y + R N u ε n φ x θ d x 1 2 s , α , θ + ε n R N ( I α * u ε n 2 s , α , θ + ε n ) u ε n 2 s , α , θ + ε n 2 u φ d x λ 2 q R N ( I α * u ε n q ) u ε n q 2 u φ d x 1 2 s , α * ε n R N ( I α * u ε n 2 s , α * ε n ) u ε n 2 s , α * ε n 2 u ε n φ d x λ 1 2 s , α , θ + ε n + 1 q R N ( I α * u ε n 2 s , α , θ + ε n ) u ε n q 2 u ε n φ d x λ 1 q + 1 2 s , α * ε n R N ( I α * u ε n q ) u ε n 2 s , α * ε n 2 u ε n φ d x 1 2 s , α , θ + ε n + 1 2 s , α * ε n R N ( I α * u ε n 2 s , α , θ + ε n ) u ε n 2 s , α * ε n 2 u ε n φ d x = R 2 N ( u 0 ( x ) u 0 ( y ) ) ( φ ( x ) φ ( y ) ) x y N + 2 s d x d y + R N u 0 φ x θ d x 1 2 s , α , θ R N ( I α * u 0 2 s , α , θ ) u 0 2 s , α , θ 2 u 0 φ d x λ 2 q R N ( I α * u 0 q ) u 0 q 2 u φ d x 1 2 s , α * R N ( I α * u 0 2 s , α * ) u 0 2 s , α * 2 u 0 φ d x λ 1 2 s , α , θ + 1 q R N ( I α * u 0 2 s , α , θ ) u 0 q 2 u 0 φ d x λ 1 q + 1 2 s , α * R N ( I α * u 0 q ) u 0 2 s , α * 2 u 0 φ d x 1 2 s , α , θ + 1 2 s , α * R N ( I α * u 0 2 s , α , θ ) u 0 2 s , α * 2 u 0 φ d x = J ( u 0 ) , φ .

This completes the proof.□

The following is the modified version of the generalized Lions-type theorem. As we mentioned earlier, it plays a key role in recovering the compactness.

Proposition 3.1

Let N 3 , s 1 2 , 1 and θ ( 0 , 2 s ) hold. Let { u n } W rad s , 2 ( R N , θ ) be a bounded sequence such that

limsup n R N u n 2 s , θ * d x > 0 a n d limsup n R N u n 2 s * d x > 0 .

Then, we have u n u 0 in L loc 2 ( R N ) .

The proof is same as in [20, Theorem 1.3]. So we omit it here for the details.

We are finally ready to prove Theorem 1.2.

Proof of Theorem 1.2

We divide the proof into three steps.

Step 1. Let { ε n } 0 and { u ε n } be a minimizing sequence of J ε n . Then

(3.21) J ε n ( u ε n ) = m ε n and P ε n ( u ε n ) = 0 as n .

We claim that { u ε n } is nonvanishing in L 2 s , θ * ( R N ) and L 2 s * ( R N ) . On the contrary, we suppose

(3.22) either lim n R N u ε n 2 s , θ * d x = 0 or lim n R N u ε n 2 s * d x = 0 .

Hence, by Proposition 1.3 and the Hölder and Young inequalities, we observe that

(3.23) R N ( I α * u ε n 2 s , α , θ + ε n ) u ε n 2 s , α , θ + ε n d x C N , α R N u ε n 2 s , θ * + 2 N ε n N + α d x N + α N C N , α R N u ε n 2 s , θ * d x ( 2 s * 2 s , θ * ) ( N + α ) 2 N ε n ( 2 s * 2 s , θ * ) N R N u ε n 2 s * d x 2 ε n 2 s * 2 s , θ * C N , α 2 s * 2 s , θ * 2 ε n 2 s * 2 s , θ * R N u ε n 2 s , θ * d x ( 2 s * 2 s , θ * ) ( N + α ) 2 N ε n ( 2 s * 2 s , θ * 2 ε n ) N + 2 ε n 2 s * 2 s , θ * R N u ε n 2 s * d x

and

(3.24) R N ( I α * u ε n q ) u ε n q d x C N , α R N u ε n 2 N q N + α d x N + α N C N , α R N u ε n 2 s , θ * d x ( N + α ) 2 s * 2 N q ( 2 s * 2 s , θ * ) N R N u ε n 2 s * d x 2 N q 2 s , θ * ( N + α ) ( 2 s * 2 s , θ * ) N

and

(3.25) R N ( I α * u ε n 2 s , α * ε n ) u ε n 2 s , α * ε n d x C N , α R N u ε n 2 s * 2 N ε n N + α d x N + α N C N , α R N u ε n 2 s , θ * d x 2 ε n 2 s * 2 s , θ * R N u ε n 2 s * d x ( 2 s * 2 s , θ * ) ( N + α ) 2 N ε n ( 2 s * 2 s , θ * ) N C N , α 2 ε n 2 s * 2 s , θ * R N u ε n 2 s , θ * d x + 2 s * 2 s , θ * 2 ε n 2 s * 2 s , θ * R N u ε n 2 s * d x ( 2 s * 2 s , θ * ) ( N + α ) 2 N ε n ( 2 s * 2 s , θ * 2 ε n ) N

and

(3.26) R N ( I α * u ε n 2 s , α , θ + ε n ) u ε n q d x C N , α R N u ε n 2 s , θ * + 2 N ε n N + α d x N + α 2 N R N u ε n 2 N q N + α d x N + α 2 N C N , α R N u ε n 2 s , θ * d x ( 2 2 s * 2 s , θ * ) ( N + α ) 2 N ( q + ε n ) 2 ( 2 s * 2 s , θ * ) N R N u ε n 2 s * d x 2 N ( q + ε n ) 2 s , θ * ( N + α ) 2 ( 2 s * 2 s , θ * ) N

and

(3.27) R N ( I α * u ε n q ) u ε n 2 s , α * ε n d x C N , α R N u ε n 2 N q N + α d x N + α 2 N R N u ε n 2 s * 2 N ε n N + α d x N + α 2 N C N , α R N u ε n 2 s , θ * d x ( N + α ) 2 s * 2 N ( q ε n ) 2 ( 2 s * 2 s , θ * ) N R N u ε n 2 s * d x ( 2 s * 2 2 s , θ * ) ( N + α ) + 2 N ( q ε n ) 2 ( 2 s * 2 s , θ * ) N

and

(3.28) R N ( I α * u ε n 2 s , α , θ + ε n ) u ε n 2 s , α * ε n d x C N , α R N u ε n 2 s , θ * + 2 N ε n N + α d x N + α 2 N R N u ε n 2 s * 2 N ε n N + α d x N + α 2 N C N , α R N u ε n 2 s , θ * d x N + α 2 N R N u ε n 2 s * d x N + α 2 N .

Gathering together (3.22)–(3.28) and P ε n ( u ε n ) = 0 , it follows that

0 = lim n N 2 s 2 u ε n D s , 2 ( R N ) 2 + N θ 2 u ε n L 2 ( R N , θ ) 2 N + α 2 ( 2 s , α , θ + ε n ) 2 R N ( I α * u ε n 2 s , α , θ + ε n ) u ε n 2 s , α , θ + ε n d x λ 2 ( N + α ) 2 q 2 R N ( I α * u ε n q ) u ε n q d x λ ( N + α ) ( 2 s , α , θ + ε n ) q R N ( I α * u ε n 2 s , α , θ + ε n ) u ε n q d x λ ( N + α ) q ( 2 s , α * ε n ) R N ( I α * u ε n q ) u ε n 2 s , α * ε n d x N + α 2 ( 2 s , α * ε n ) 2 R N ( I α * u ε n 2 s , α * ε n ) u ε n 2 s , α * ε n d x N + α ( 2 s , α , θ + ε n ) ( 2 s , α * ε n ) R N ( I α * u ε n 2 s , α , θ + ε n ) u ε n 2 s , α * ε n d x = lim n N 2 s 2 u ε n D s , 2 ( R N ) 2 + N θ 2 u ε n L 2 ( R N , θ ) 2 .

It leads to a contradiction with (3.21). As a consequence of (3.21), we can notice { u ε n } is bounded in W rad s , 2 ( R N , θ ) . By virtue of Proposition 3.1, we deduce equation ( C ε ) possesses a nonnegative ground state solution u ε n .

Step 2. Up to subsequences, we assume that there exists u 0 such that

u ε n u 0 in W rad s , 2 ( R N , θ ) , u ε n u 0 in L r ( R N ) for r ( 2 s , θ * , 2 s * ) and u ε n u 0 a.e.in R N .

We deduce from Lemma 3.6 that u 0 is a weak solution of equation ( C ). In the following, we shall show u 0 0 . Recalling Lemma 3.5, then

u ε n W rad s , 2 ( R N , θ ) C .

In what follows, we distinguish three cases:

Case 1. R N u ε n 2 s , θ * d x 0 and R N u ε n 2 s * d x C .

Case 2. R N u ε n 2 s , θ * d x C and R N u ε n 2 s * d x 0 .

Case 3. R N u ε n 2 s , θ * d x C and R N u ε n 2 s * d x C .

If Case 1 happens, exploiting (1.7) and (3.23)–(3.28), we see that

N 2 s 2 u ε n D s , 2 ( R N ) 2 N + α 2 ( 2 s , α , θ + ε n ) 2 R N ( I α * u ε n 2 s , α , θ + ε n ) u ε n 2 s , α , θ + ε n d x + λ 2 ( N + α ) 2 q 2 R N ( I α * u ε n q ) u ε n q d x + N + α 2 ( 2 s , α * ε n ) 2 R N ( I α * u ε n 2 s , α * ε n ) u 2 s , α * ε n d x + λ ( N + α ) ( 2 s , α , θ + ε n ) q R N ( I α * u ε n 2 s , α , θ + ε n ) u ε n q d x + λ ( N + α ) q ( 2 s , α * ε n ) R N ( I α * u ε n q ) u ε n 2 s , α * ε n d x + N + α ( 2 s , α , θ + ε n ) ( 2 s , α * ε n ) R N ( I α * u ε n 2 s , α , θ + ε n ) u ε n 2 s , α * ε n d x C N , α ( N + α ) ( 2 s * 2 s , θ * 2 ε n ) 2 ( 2 s , α * ε n ) 2 ( 2 s * 2 s , θ * ) R N u ε n 2 s * d x ( 2 s * 2 s , θ * ) ( N + α ) 2 N ε n ( 2 s * 2 s , θ * 2 ε n ) N C N , α ( N + α ) ( 2 s * 2 s , θ * 2 ε n ) 2 ( 2 s , α * ε n ) 2 ( 2 s * 2 s , θ * ) 1 S s 2 s * [ ( 2 s * 2 s , θ * ) ( N + α ) 2 N ε n ] 2 ( 2 s * 2 s , θ * 2 ε n ) N × u ε n D s , 2 ( R N ) 2 s * [ ( 2 s * 2 s , θ * ) ( N + α ) 2 N ε n ] ( 2 s * 2 s , θ * 2 ε n ) N .

From this, we obtain

u ε n D s , 2 ( R N ) 2 ( N 2 s ) ( 2 s , α * ε n ) 2 ( 2 s * 2 s , θ * ) C N , α ( N + α ) ( 2 s * 2 s , θ * 2 ε n ) 2 ( 2 s * 2 s , θ * 2 ε n ) N 2 s * [ ( 2 s * 2 s , θ * ) ( N + α ) 2 N ε n ] 2 ( 2 s * 2 s , θ * 2 ε n ) N × S s 2 s * [ ( 2 s * 2 s , θ * ) ( N + α ) 2 N ε n ] 2 s * [ ( 2 s * 2 s , θ * ) ( N + α ) 2 N ε n ] 2 ( 2 s * 2 s , θ * 2 ε n ) N .

Combining the latter formula and Lemma 3.5 (ii), we immediately obtain

m lim n m ε n = lim n J ε n ( u ε n ) 1 N + α P ε n ( u ε n ) lim n α + 2 s 2 ( N + α ) u ε n D s , 2 ( R N ) 2 α + 2 s 2 ( N + α ) ( 2 s , α * ) 2 ( N 2 s ) C N , α ( N + α ) 2 N 2 s * ( N + α ) 2 N S s 2 s * ( N + α ) 2 s * ( N + α ) 2 N ,

which yields a contradiction with Lemma 3.3. So, Case 1 does not happen.

If Case 2 happens, combining (1.8) and (3.23)–(3.28), we obtain

N 2 s 2 u ε n W rad s , 2 ( R N , θ ) 2 N + α 2 ( 2 s , α , θ + ε n ) 2 R N ( I α * u ε n 2 s , α , θ + ε n ) u ε n 2 s , α , θ + ε n d x + λ 2 ( N + α ) 2 q 2 R N ( I α * u ε n q ) u ε n q d x + N + α 2 ( 2 s , α * ε n ) 2 R N ( I α * u ε n 2 s , α * ε n ) u 2 s , α * ε n d x + λ ( N + α ) ( 2 s , α , θ + ε n ) q R N ( I α * u ε n 2 s , α , θ + ε n ) u ε n q d x + λ ( N + α ) q ( 2 s , α * ε n ) R N ( I α * u ε n q ) u ε n 2 s , α * ε n d x + N + α ( 2 s , α , θ + ε n ) ( 2 s , α * ε n ) R N ( I α * u ε n 2 s , α , θ + ε n ) u ε n 2 s , α * ε n d x C N , α ( N + α ) ( 2 s * 2 s , θ * 2 ε n ) 2 ( 2 s , α , θ + ε n ) 2 ( 2 s * 2 s , θ * ) R N u ε n 2 s , θ * d x ( 2 s * 2 s , θ * ) ( N + α ) 2 N ε n ( 2 s * 2 s , θ * 2 ε n ) N C N , α ( N + α ) ( 2 s * 2 s , θ * 2 ε n ) 2 ( 2 s , α , θ + ε n ) 2 ( 2 s * 2 s , θ * ) 1 S s , θ 2 s , θ * [ ( 2 s * 2 s , θ * ) ( N + α ) 2 N ε n ] 2 ( 2 s * 2 s , θ * 2 ε n ) N × u ε n W rad s , 2 ( R N , θ ) 2 s , θ * [ ( 2 s * 2 s , θ * ) ( N + α ) 2 N ε n ] ( 2 s * 2 s , θ * 2 ε n ) N ,

which implies

u ε n W rad s , 2 ( R N , θ ) 2 ( N 2 s ) ( 2 s , α , θ + ε n ) 2 ( 2 s * 2 s , θ * ) C N , α ( N + α ) ( 2 s * 2 s , θ * 2 ε n ) 2 ( 2 s * 2 s , θ * 2 ε n ) N 2 s , θ * [ ( 2 s * 2 s , θ * ) ( N + α ) 2 N ε n ] 2 ( 2 s * 2 s , θ * 2 ε n ) N × S s , θ 2 s , θ * [ ( 2 s * 2 s , θ * ) ( N + α ) 2 N ε n ] 2 s , θ * [ ( 2 s * 2 s , θ * ) ( N + α ) 2 N ε n ] 2 ( 2 s * 2 s , θ * 2 ε n ) N .

We obtain from this and Lemma 3.5 (ii) that

m lim n m ε n = lim n J ε n ( u ε n ) 1 N + α P ε n ( u ε n ) lim n α + 2 s 2 ( N + α ) u ε n W rad s , 2 ( R N , θ ) 2 α + 2 s 2 ( N + α ) ( 2 s , α , θ ) 2 ( N 2 s ) C N , α ( N + α ) 2 N 2 s , θ * ( N + α ) 2 N S s , θ 2 s , θ * ( N + α ) 2 s , θ * ( N + α ) 2 N .

It gives a contradiction with Lemma 3.3. Hence, Case 2 does not happen.

Hence, we observe that only Case 3 happen. From Proposition 3.1, one has u 0 0 in L loc 2 ( R N ) . Thus, u 0 is a nontrivial solution of equation ( C ).

Step 3. Finally, we claim u 0 is a ground state solution of equation ( C ). Combining Lemmas 3.5,3.6, and the Brézis-Lieb lemma [11], then

m J ( u 0 ) = J ( u 0 ) 1 N + α P ( u 0 ) lim n J ε n ( u ε n ) 1 N + α P ε n ( u ε n ) m .

So, we have J ( u 0 ) = m . Furthermore, we can choose u 0 0 . This ends the proof.□

  1. Funding information: J. Yang is supported by Natural Science Foundation of Hunan Province of China (Nos. 2023JJ30482 and 2022JJ30463), Research Foundation of Education Bureau of Hunan Province (22A0540).

  2. Conflict of interest: The authors declare that they have no conflicts of interest to this work.

  3. Data availability statement: Data sharing is not applicable to this article as no new data were created or analyzed in this study.

References

[1] C. Alves, A. Nóbrega, and M. Yang, Multi-bump solutions for Choquard equation with deepening potential well, Calc. Var. Partial Differential Equations 55 (2016), no. 3, Art. 48, 28. 10.1007/s00526-016-0984-9Search in Google Scholar

[2] V. Ambrosio, Multiplicity and concentration results for a fractional Choquard equation via penalization method, Potential Anal. 50 (2019), no. 1, 55–82. 10.1007/s11118-017-9673-3Search in Google Scholar

[3] M. Badiale and S. Rolando, A note on nonlinear elliptic problems with singular potentials, Atti Accad. Naz. Lincei Rend. Lincei Mat. Appl. 17 (2006), no. 1, 1–13. 10.4171/rlm/450Search in Google Scholar

[4] M. Badiale, V. Benci, and S. Rolando, A nonlinear elliptic equation with singular potential and applications to nonlinear field equations, J. Eur. Math. Soc. (JEMS) 9 (2007), no. 3, 355–381. 10.4171/jems/83Search in Google Scholar

[5] P. Belchior, H. Bueno, O. Miyagaki, and G. Pereira, Remarks about a fractional Choquard equation: ground state, regularity and polynomial decay, Nonlinear Anal. 164 (2017), 38–53. 10.1016/j.na.2017.08.005Search in Google Scholar

[6] J. Bellazzini, M. Ghimenti, C. Mercuri, V. Moroz, and J. Van Schaftingen, Sharp Gagliardo-Nirenberg inequalities in fractional Coulomb-Sobolev spaces, Trans. Amer. Math. Soc. 370 (2018), no. 11, 8285–8310. 10.1090/tran/7426Search in Google Scholar

[7] V. Benci and D. Fortunato, Variational methods in nonlinear field equations, Springer Monographs in Mathematics, Springer, Cham, 2014, Solitary waves, hylomorphic solitons and vortices. 10.1007/978-3-319-06914-2Search in Google Scholar

[8] H. Berestycki and P. Lions, Nonlinear scalar field equations. I. Existence of a ground state, Arch. Rational Mech. Anal. 82 (1983), no. 4, 313–345. 10.1007/BF00250555Search in Google Scholar

[9] S. Bhattarai, On fractional Schrödinger systems of Choquard type, J. Differential Equations 263 (2017), no. 6, 3197–3229. 10.1016/j.jde.2017.04.034Search in Google Scholar

[10] B. Bieganowski and S. Secchi, The semirelativistic Choquard equation with a local nonlinear term, Discrete Contin. Dyn. Syst. 39 (2019), no. 7, 4279–4302. 10.3934/dcds.2019173Search in Google Scholar

[11] H. Brézis and E. Lieb, A relation between pointwise convergence of functions and convergence of functionals, Proc. Amer. Math. Soc. 88 (1983), no. 3, 486–490. 10.1090/S0002-9939-1983-0699419-3Search in Google Scholar

[12] L. Caffarelli, Surfaces minimizing nonlocal energies, Atti Accad. Naz. Lincei Rend. Lincei Mat. Appl. 20 (2009), no. 3, 281–299. 10.4171/rlm/547Search in Google Scholar

[13] D. Cassani, J. Van Schaftingen, and J. Zhang, Groundstates for Choquard type equations withHardy-Littlewood-Sobolev lower critical exponent, Proc. Roy. Soc. Edinburgh Sect. A 150 (2020), no. 3, 1377–1400. 10.1017/prm.2018.135Search in Google Scholar

[14] S. Cingolani, M. Clapp, and S. Secchi, Multiple solutions to a magnetic nonlinear Choquard equation, Z. Angew. Math. Phys. 63 (2012), no. 2, 233–248. 10.1007/s00033-011-0166-8Search in Google Scholar

[15] S. Cingolani and K. Tanaka, Semi-classical states for the nonlinear Choquard equations: existence, multiplicity and concentration at a potential well, Rev. Mat. Iberoam. 35 (2019), no. 6, 1885–1924. 10.4171/rmi/1105Search in Google Scholar

[16] M. Conti, S. Crotti, and D. Pardo, On the existence of positive solutions for a class of singular elliptic equations, Adv. Differential Equations 3 (1998), no. 1, 111–132. 10.57262/ade/1366399907Search in Google Scholar

[17] P. D’Avenia, G. Siciliano, and M. Squassina, On fractional Choquard equations, Math. Models Methods Appl. Sci. 25 (2015), no. 8, 1447–1476. 10.1142/S0218202515500384Search in Google Scholar

[18] E. Di Nezza, G. Palatucci, and E. Valdinoci, Hitchhiker’s guide to the fractional Sobolev spaces, Bull. Sci. Math. 136 (2012), no. 5, 521–573. 10.1016/j.bulsci.2011.12.004Search in Google Scholar

[19] S. Dipierro, G. Palatucci, and E. Valdinoci, Existence and symmetry results for a Schrödinger type problem involving the fractional Laplacian, Matematiche (Catania) 68 (2013), no. 1, 201–216. Search in Google Scholar

[20] Z. Feng and Y. Su, Lions-type theorem of the fractional Laplacian and applications, Dyn. Partial Differential Equations 18 (2021), no. 3, 211–230. 10.4310/DPDE.2021.v18.n3.a3Search in Google Scholar

[21] X. He and V. D. Rădulescu, Small linear perturbations of fractional Choquard equations with critical exponent, J. Differential Equations 282 (2021), 481–540. 10.1016/j.jde.2021.02.017Search in Google Scholar

[22] N. Laskin, Fractional quantum mechanics and Lévy path integrals, Phys. Lett. A 268 (2000), no. 4–6, 298–305. 10.1016/S0375-9601(00)00201-2Search in Google Scholar

[23] N. Laskin, Fractional Schrödinger equation, Phys. Rev. E (3) 66 (2002), no. 5, 056108, 7. 10.1103/PhysRevE.66.056108Search in Google Scholar PubMed

[24] E. Lieb, Existence and uniqueness of the minimizing solution of Choquardas nonlinear equation, Studies Appl. Math. 57 (1976/77), no. 2, 93–105. 10.1002/sapm197757293Search in Google Scholar

[25] E. Lieb and M. Loss, Analysis, Graduate Studies in Mathematics, vol. 14, American Mathematical Society, Providence, RI, 1997. Search in Google Scholar

[26] P. Lions, The Choquard equation and related questions, Nonlinear Anal. 4 (1980), no. 6, 1063–1072. 10.1016/0362-546X(80)90016-4Search in Google Scholar

[27] Z. Liu, H. Luo, and J. Zhang, Existence and multiplicity of bound state solutions to a Kirchhoff type equation with a general nonlinearity, J. Geom. Anal, 32 (2022), no. 4, Paper No. 125, 25pp. 10.1007/s12220-021-00849-0Search in Google Scholar

[28] Z. Liu, V. Rădulescu, C. Tang, and J. Zhang, Another look at planar Schrödinger-Newton systems, J. Differential Equations 328 (2022), 65–104. 10.1016/j.jde.2022.04.035Search in Google Scholar

[29] Z. Liu, V. Rădulescu, and Z. Yuan, Concentration of solutions for fractional Kirchhoff equations with discontinuous reaction, Z. Angew. Math. Phys 73 (2022), no. 5, Paper No. 211, 23. 10.1007/s00033-022-01849-ySearch in Google Scholar

[30] Z. Liu, V. Rădulescu, and J. Zhang, A planar Schrödinger-Newton system with Trudinger-Moser critical growth, Calc. Var. Partial Differential Equations, 62 (2023), no. 4, 111122. 10.1007/s00526-023-02463-0Search in Google Scholar

[31] P. Menzala, On regular solutions of a nonlinear equation of Choquard’s type, Proc. Roy. Soc. Edinburgh Sect. A 86 (1980), no. 3–4, 291–301. 10.1017/S0308210500012191Search in Google Scholar

[32] V. Moroz and J. Van Schaftingen, Groundstates of nonlinear Choquard equations: existence, qualitative properties and decay asymptotics, J. Funct. Anal. 265 (2013), no. 2, 153–184. 10.1016/j.jfa.2013.04.007Search in Google Scholar

[33] V. Moroz and J. Van Schaftingen, Existence of groundstates for a class of nonlinear Choquardequations, Trans. Amer. Math. Soc. 367 (2015), no. 9, 6557–6579. 10.1090/S0002-9947-2014-06289-2Search in Google Scholar

[34] V. Moroz and J. Van Schaftingen, A guide to the Choquard equation, J. Fixed Point Theory Appl. 19 (2017), no. 1, 773–813. 10.1007/s11784-016-0373-1Search in Google Scholar

[35] S. Pekar, Untersuchung ber die elektronentheorie der kristalle, Akademie Verlag, Berlin, 1954. 10.1515/9783112649305Search in Google Scholar

[36] D. Ruiz and J. Van Schaftingen, Odd symmetry of least energy nodal solutions for the Choquard equation, J. Differential Equations 264 (2018), no. 2, 1231–1262. 10.1016/j.jde.2017.09.034Search in Google Scholar

[37] L. Silvestre, Regularity of the obstacle problem for a fractional power ofthe Laplace operator, Comm. Pure Appl. Math. 60 (2007), no. 1, 67–112. 10.1002/cpa.20153Search in Google Scholar

[38] A. Strauss, Existence of solitary waves in higher dimensions, Comm. Math. Phys. 55 (1977), no. 2, 149–162. 10.1007/BF01626517Search in Google Scholar

[39] J. Su, Z. Wang, and M. Willem, Nonlinear Schrödinger equations with unbounded and decaying radial potentials, Commun. Contemp. Math. 9 (2007), no. 4, 571–583. 10.1142/S021919970700254XSearch in Google Scholar

[40] S. Terracini, On positive entire solutions to a class of equations with a singular coefficient and critical exponent, Adv. Differential Equations 1 (1996), no. 2, 241–264. 10.57262/ade/1366896239Search in Google Scholar

[41] J. Wei and M. Winter, Strongly interacting bumps for the Schrödinger-Newton equations, J. Math. Phys. 50 (2009), no. 1, 012905, 22. 10.1063/1.3060169Search in Google Scholar

[42] M. Willem, Minimax theorems, Progress in Nonlinear Differential Equations and their Applications, vol. 24, Birkhäuser Boston, Inc., Boston, MA, 1996. Search in Google Scholar

[43] M. Willem, Functional analysis, Cornerstones, Birkhäuser/Springer, New York, 2013, Fundamentals and applications. 10.1007/978-1-4614-7004-5Search in Google Scholar

[44] X. Ros-Oton and S. Joaquim, The Dirichlet problem for the fractional Laplacian: regularity up to the boundary, J. Math. Pures Appl. 101 (2014), no. 3, 275–302. 10.1016/j.matpur.2013.06.003Search in Google Scholar

[45] M. Yang, W. Ye, and S. Zhao, Existence of concentrating solutions of the Hartree type Brezis-Nirenberg problem, J. Differential Equations 344 (2023), 260–324. 10.1016/j.jde.2022.10.041Search in Google Scholar

[46] M. Yang, V. Rădulescu, and X. Zhou, Critical Stein-Weiss elliptic systems: symmetry,regularity and asymptotic properties of solutions, Calc. Var. Partial Differential Equations 61 (2022), no. 3, Paper No. 109, 38pp. 10.1007/s00526-022-02221-8Search in Google Scholar

Received: 2023-04-26
Revised: 2023-06-09
Accepted: 2024-03-01
Published Online: 2024-05-09

© 2024 the author(s), published by De Gruyter

This work is licensed under the Creative Commons Attribution 4.0 International License.

Articles in the same Issue

  1. Regular Articles
  2. Existence of normalized peak solutions for a coupled nonlinear Schrödinger system
  3. Orbital stability of the trains of peaked solitary waves for the modified Camassa-Holm-Novikov equation
  4. Global boundedness in a two-dimensional chemotaxis system with nonlinear diffusion and singular sensitivity
  5. Existence and uniqueness of solution for a singular elliptic differential equation
  6. The bounded variation capacity and Sobolev-type inequalities on Dirichlet spaces
  7. Vanishing and blow-up solutions to a class of nonlinear complex differential equations near the singular point
  8. Existence, uniqueness, localization and minimization property of positive solutions for non-local problems involving discontinuous Kirchhoff functions
  9. Concentration phenomena for a fractional relativistic Schrödinger equation with critical growth
  10. Beyond the classical strong maximum principle: Sign-changing forcing term and flat solutions
  11. Monotonicity of solutions for parabolic equations involving nonlocal Monge-Ampère operator
  12. On a nonlinear Robin problem with an absorption term on the boundary and L1 data
  13. Multiple solutions for the quasilinear Choquard equation with Berestycki-Lions-type nonlinearities
  14. Gradient estimates for a class of higher-order elliptic equations of p-growth over a nonsmooth domain
  15. Global existence and decay estimates of the classical solution to the compressible Navier-Stokes-Smoluchowski equations in ℝ3
  16. Global existence and finite-time blowup for a mixed pseudo-parabolic r(x)-Laplacian equation
  17. Multiple positive solutions for a class of concave-convex Schrödinger-Poisson-Slater equations with critical exponent
  18. A minimization problem with free boundary for p-Laplacian weakly coupled system
  19. Multiplicity of semiclassical solutions for a class of nonlinear Hamiltonian elliptic system
  20. k-convex solutions for multiparameter Dirichlet systems with k-Hessian operator and Lane-Emden type nonlinearities
  21. Infinitely many solutions for Hamiltonian system with critical growth
  22. On optimal control in a nonlinear interface problem described by hemivariational inequalities
  23. Variational–hemivariational system for contaminant convection–reaction–diffusion model of recovered fracturing fluid
  24. Potential and monotone homeomorphisms in Banach spaces
  25. Critical fractional Schrödinger-Poisson systems with lower perturbations: the existence and concentration behavior of ground state solutions
  26. Supercritical Hénon-type equation with a forcing term
  27. Concentration of blow-up solutions for the Gross-Pitaveskii equation
  28. Mountain-pass-type solutions for Schrödinger equations in R2 with unbounded or vanishing potentials and critical exponential growth nonlinearities
  29. Regularity for critical fractional Choquard equation with singular potential and its applications
  30. Double phase anisotropic variational problems involving critical growth
  31. Normalized solutions of NLS equations with mixed nonlocal nonlinearities
  32. Nontrivial solutions for resonance quasilinear elliptic systems
  33. Standing waves for Choquard equation with noncritical rotation
  34. Low regularity conservation laws for Fokas-Lenells equation and Camassa-Holm equation
  35. Modified quasilinear equations with strongly singular and critical exponential nonlinearity
  36. Limit profiles and the existence of bound-states in exterior domains for fractional Choquard equations with critical exponent
  37. Nests of limit cycles in quadratic systems
  38. Regularity of minimizers for double phase functionals of borderline case with variable exponents
  39. Boundedness, existence and uniqueness results for coupled gradient dependent elliptic systems with nonlinear boundary condition
  40. Ground state solutions for magnetic Schrödinger equations with polynomial growth
  41. Nonoccurrence of Lavrentiev gap for a class of functionals with nonstandard growth
  42. Dynamics for wave equations connected in parallel with nonlinear localized damping
  43. Boussinesq's equation for water waves: Asymptotics in Sector I
  44. Optimal global second-order regularity and improved integrability for parabolic equations with variable growth
  45. Normalized solutions of Schrödinger equations involving Moser-Trudinger critical growth
  46. Normalized solutions for the double-phase problem with nonlocal reaction
  47. Normalized solutions for Sobolev critical fractional Schrödinger equation
  48. Well-posedness of Cauchy problem of fractional drift diffusion system in non-critical spaces with power-law nonlinearity
  49. Improved results on planar Klein-Gordon-Maxwell system with critical exponential growth
  50. Existence and multiplicity of solutions for a new p(x)-Kirchhoff equation
  51. Quasiconvex bulk and surface energies: C1,α regularity
  52. Time decay estimates of solutions to a two-phase flow model in the whole space
  53. Bounds for the sum of the first k-eigenvalues of Dirichlet problem with logarithmic order of Klein-Gordon operators
  54. The properties of a new fractional g-Laplacian Monge-Ampère operator and its applications
  55. A fractional profile decomposition and its application to Kirchhoff-type fractional problems with prescribed mass
  56. Cauchy problem for a non-Newtonian filtration equation with slowly decaying volumetric moisture content
  57. Multiplicity of normalized semi-classical states for a class of nonlinear Choquard equations
  58. The second Yamabe invariant with boundary
  59. Peaked solitary waves and shock waves of the Degasperis-Procesi-Kadomtsev-Petviashvili equation
  60. Normalized solutions for the Choquard equations with critical nonlinearities
  61. Boundedness and long-time behavior in a parabolic-elliptic system arising from biological transport networks
  62. Initial boundary value problem and exponential stability for the planar magnetohydrodynamics equations with temperature-dependent viscosity
  63. Nonexistence of mean curvature flow solitons with polynomial volume growth immersed in certain semi-Riemannian warped products
  64. Analysis of a vector-borne disease model with vector-bias mechanism in advective heterogeneous environment
  65. Normalized solutions for the Kirchhoff equation with combined nonlinearities in ℝ4
  66. Nonexistence of the compressible Euler equations with space-dependent damping in high dimensions
  67. Multiplicity of solutions for a nonhomogeneous quasilinear elliptic equation with concave-convex nonlinearities
  68. On semilinear inequalities involving the Dunkl Laplacian and an inverse-square potential outside a ball
  69. Besov regularity for the elliptic p-harmonic equations in the non-quadratic case
  70. Uniqueness and nondegeneracy of ground states for Δ u + u = ( I α u 2 ) u in R 3 when α is close to 2
  71. Existence of solutions for a class of quasilinear Schrödinger equations with Choquard-type nonlinearity
  72. Weighted Hardy-Adams inequality on unit ball of any even dimension
  73. Existence and regularity for a p-Laplacian problem in ℝN with singular, convective, and critical reaction
  74. Temporal periodic solutions of non-isentropic compressible Euler equations with geometric effects
  75. Nodal solutions for a zero-mass Chern-Simons-Schrödinger equation
  76. The modified quasi-boundary-value method for an ill-posed generalized elliptic problem
  77. Nonlocal heat equations with generalized fractional Laplacian
  78. Choquard equations with recurrent potentials
  79. Local and global well-posedness of the Maxwell-Bloch system of equations with inhomogeneous broadening
  80. The Riemann problem for two-layer shallow water equations with bottom topography
  81. Global stability and asymptotic profiles of a partially degenerate reaction diffusion Cholera model with asymptomatic individuals
  82. On Kirchhoff-Schrödinger-Poisson-type systems with singular and critical nonlinearity
  83. Energy-variational solutions for viscoelastic fluid models
  84. Stability on 3D Boussinesq system with mixed partial dissipation
  85. Existence and multiplicity results for non-autonomous second-order Hamiltonian systems
  86. Erratum
  87. Erratum to “Normalized solutions for the Kirchhoff equation with combined nonlinearities in ℝ4
Downloaded on 7.9.2025 from https://www.degruyterbrill.com/document/doi/10.1515/anona-2024-0001/html
Scroll to top button