Home Well-posedness of Cauchy problem of fractional drift diffusion system in non-critical spaces with power-law nonlinearity
Article Open Access

Well-posedness of Cauchy problem of fractional drift diffusion system in non-critical spaces with power-law nonlinearity

  • Caihong Gu and Yanbin Tang ORCID logo EMAIL logo
Published/Copyright: August 5, 2024

Abstract

In this article, we consider the global and local well-posedness of the mild solutions to the Cauchy problem of fractional drift diffusion system with higher-order nonlinearity. The main difficulty comes from the higher-order nonlinearity. Instead of the convention that people always focus on the properties of the solution in critical spaces, here we are interested in non-critical spaces such as supercritical Sobolev spaces and subcritical Lebesgue spaces. For the initial data in these non-critical spaces, using the properties of fractional heat semigroup and the classical Hardy-Littlewood-Sobolev inequality, we obtain the existence and uniqueness of the mild solution, together with the decaying rate estimates in terms of time variable.

MSC 2010: 35K45; 35K55; 35Q60; 35B40

1 Introduction

In this article, we consider the well-posedness to the Cauchy problem of fractional drift diffusion system with higher-order nonlinearity

(1.1) t v + Λ α v = ( v m ϕ ) , t > 0 , x R n , t w + Λ α w = ( w m ϕ ) , t > 0 , x R n , Δ ϕ = v w , t > 0 , x R n , v ( x , 0 ) = v 0 ( x ) , w ( x , 0 ) = w 0 ( x ) , x R n ,

where m 1 is an integer, v ( x , t ) and w ( x , t ) are the densities of negatively and positively charged particles and ϕ ( x , t ) is the electric potential determined by the Poisson equation Δ ϕ = v w . Λ = Δ is the Calderón-Zygmund operator [1]. The difficulties mainly come from higher-order nonlinear couplings.

In the physical literature, such fractal anomalous diffusions have been recently enthusiastically embraced by a slew of investigators in the context of hydrodynamics, acoustics, trapping effects in surface diffusion, statistical mechanics, relaxation phenomena, and biology. For instance, astrophysics is a source of mean-field models of gravitationally attracting particles going back to the famous Chandrasekhar equation for the equilibrium of radiating stars [2,3]. Another source of related models is mathematical biology where chemotaxis (haptotaxis, angiogenesis, etc.) phenomena for populations of either cells or (micro)organisms are described by various modifications of the Keller-Segel systems [415].

In our previous study [16], we considered the global existence, regularizing decay rate, and asymptotic behavior of mild solutions to the Cauchy problem of fractional drift diffusion system [1723] with power-law nonlinearity, we only studied the problem in the critical Besov spaces, and we obtained the global well-posedness since the critical index provides the minimal regularity for the initial data to ensure the existence of the mild solutions. But in this article, we are interested in the Cauchy problem of the drift-diffusion equation in non-critical spaces such as supercritical Sobolev spaces and subcritical Lebesgue spaces. Using the properties of fractional heat semigroup and the classical estimates of fractional heat kernel, we first proved the global-in-time existence and uniqueness of the mild solutions in the frame of mixed time-space Besov space with multi-linear continuous mappings. Then, we showed the asymptotic behavior and regularizing-decay rate estimates of the solution to equations with power-law nonlinearity by the method of multi-linear operator and the classical Hardy-Littlewood-Sobolev inequality.

But now we are interested in the well-posedness of solution to the Cauchy problem (1.1) with the initial data in some non-critical spaces, such as local solutions in supercritical Sobolev spaces H p s ( R n ) and global solutions in subcritical Lebesgue spaces L p ( R n ) .

By the fundamental solution of Laplacian

(1.2) Φ n ( x ) = 1 2 x , n = 1 , 1 2 π ln x , n = 2 , 1 n ( n 2 ) ω ( n ) x n 2 , n 3 ,

where ω ( n ) denotes the volume of the unit ball in R n , the electric potential ϕ can be expressed by the convolution

ϕ = ( Δ ) 1 ( w v ) = Φ n * ( w v ) = R n Φ n ( x y ) ( w v ) ( y ) d y .

Λ = Δ is the Calderón-Zygmund operator, and the fractional Laplacian Λ α = ( Δ ) α 2 with 1 < α < 2 n is a non-local fractional differential operator defined as

Λ α v ( x ) = 1 ξ α v ( ξ ) ,

where and 1 are the Fourier transform and its inverse [1].

In probabilistic terms, replacing the Laplacian Δ by its fractional power Λ α = ( Δ ) α 2 , it leads to interesting and largely open questions of extensions of results for Brownian motion-driven stochastic equations to those driven by Lévy α -stable flights [24].

An important technical difficulty is that the densities of the semigroups generated by Λ α = ( Δ ) α 2 do not decay rapidly in x R n as is the case of the heat semigroup S ( t ) = e t Δ ( α = 2 ) , the Gauss-Weierstrass kernel K t ( x ) = 1 ( e t ξ 2 ) decays exponentially, while the densities 1 ( e t ξ α ) ( 0 < α < 2 ) of non-Gaussian Lévy α -stable semigroups S α ( t ) = e t ( Δ ) α 2 have only an algebraic decay rate x n α . For the fractional operator and the fractional heat kernel, we can also refer [2531].

By the fractional heat semigroup e t Λ α and the well-known Duhamel principle, we rewrite System (1.1) as a system of integral equations

(1.3) v ( t ) = e t Λ α v 0 + B ( v , , v , w ) , w ( t ) = e t Λ α w 0 + B ( w , , w , v ) ,

(1.4) B ( v , , v m , w ) = 0 t e ( t τ ) Λ α ( v m ( Δ ) 1 ( w v ) ) ( τ ) d τ .

A solution of (1.3) and (1.4) is called a mild solution of the Cauchy problem (1.1).

As usual, the fractional Sobolev spaces and their homogeneous versions are defined by

H p s ( R n ) = { f S ( R n ) : Λ s f L p } , H ˙ p s ( R n ) = { f S ( R n ) : Λ ˙ s f L p } ,

where Λ s and Λ ˙ s are the operators with symbols Λ s ( ξ ) = ( 1 + ξ 2 ) s 2 and Λ ˙ s = ξ s (these spaces are sometimes also denoted L p , s ( R n ) , see [32]).

If ( v ( x , t ) , w ( x , t ) ) is a solution of the Cauchy problem (1.1), then for any λ > 0 ,

( v λ ( x , t ) , w λ ( x , t ) ) = λ α m v ( λ x , λ α t ) , λ α m w ( λ x , λ α t )

is also a solution of the Cauchy problem (1.1) with the initial data

( v λ ( x , 0 ) , w λ ( x , 0 ) ) = λ α m v 0 ( λ x ) , λ α m w 0 ( λ x ) .

( v λ ( x , t ) , w λ ( x , t ) ) is called the self-similar solution. We can verify that H p s c ( R n ) ( s c = n p α m ) is a critical space for the Cauchy problem (1.1), i.e., the self-similar initial data ( v 0 ( x ) , w 0 ( x ) ) are invariant under the norm H p s c , thus the solution ( v λ ( x , t ) , w λ ( x , t ) ) of System (1.1) too. This means that the index s c provides the minimal regularity for the initial data to ensure the well-posedness of the Cauchy problem (1.1).

To obtain the local existence of the solutions in supercritical Sobolev spaces, we need the following two assumptions on s :

(H1) There exists m > 1 such that

s m n p ( m + 1 ) 1 m + 1 .

(H2) There exists m > 1 such that

m n p ( m + 1 ) 1 m + 1 < s < min m n p ( m + 1 ) + 1 , n p 1 .

To avoid technical problems, we will assume that

(1.5) s n p n + 1 m + 1

and that

(1.6) s 0 .

We want to solve the Cauchy problem (1.1) in supercritical spaces H p s ( R n ) ; the main idea is to counterbalance the loss of smoothness coming from the nonlinear terms by the smoothing effects of the heat kernel. To measure the loss of smoothness on the H p s ( R n ) scale coming from the composition by F ( v , w ) = v m ( Δ ) 1 ( v w ) , we give the following theorem.

Theorem 1.1

Let p ( 0 , ) and s satisfy that

(1.7) max 0 , n p n + 1 m + 1 < s < n p 1 .

Denote s m = ( m + 1 ) s m n p + 1 . If (H1) or(H2) is fulfilled, then for all ( v , w ) H p s ( R n ) , F ( v , w ) H p s m ( R n ) . Furthermore, there exists a constant C independent of ( v , w ) such that

F ( v , w ) H p s m C v H p s m v w H p s .

Remark 1.1

  1. Note that the condition s m n p ( m + 1 ) 1 m + 1 in assumption (H1) is equivalent to s m 0 . In the same way, the condition m n p ( m + 1 ) 1 m + 1 < s < min m n p ( m + 1 ) + 1 , n p 1 in assumption (H2) is equivalent to 0 < s m < m .

  2. The hypothesis s > max 0 , n p n + 1 m + 1 ensures that F ( v , w ) is well defined as an element of tempered distribution space.

  3. The value of s m given by Theorem 1.1 is optimal. To see this, we have just to consider the example of v ( x ) = w ( x ) = ψ ( x ) x β , where ψ is a truncation function near 0, and β is an arbitrary nonnegative constant.

Using the nonlinear estimates given by Theorem 1.1 and the fixed point theorem, we prove the following result about the local Cauchy problem in supercritical Sobolev spaces H p s ( R n ) .

Theorem 1.2

(Local existence in H p s ) Let p ( 1 , ) . Assume that (1.5) and (1.6) hold, and(H1) or(H2) is fulfilled.

  1. If the initial data ( v 0 , w 0 ) H p s ( R n ) with s > s c = n p α m , there exists a unique solution ( V ( t , x ) , W ( t , x ) ) of (1.1) in C ( [ 0 , T m ) , H p s ) with T m C 8 ( v 0 , w 0 ) H p s ν with ν = s s c α . If T m < + , then

    lim t T m ( V ( t , x ) , W ( t , x ) ) H p s = + .

  2. Furthermore, we have the following smoothing effect: B ( V , W ) C ( [ 0 , T m ) , H p s + θ ) for all θ < m ( s s c ) if s < n p 1 .

  3. Let ( V , W ) C ( [ 0 , T 1 ) , H p s ) and ( V ˜ , W ˜ ) C ( [ 0 , T 2 ) , H p s ) be two solutions of (1.1) for the respective initial data ( v 0 , w 0 ) and ( v ˜ 0 , w ˜ 0 ) . Then, for all T < min { T 1 , T 2 } ,

    ( V V ˜ , W W ˜ ) C ( [ 0 , T ) , H p s ) C ( T ) ( v 0 v ˜ 0 H p s 1 2 + w 0 w ˜ 0 H p s 1 2 ) .

Because of the definition of s c , we see that L p ( R n ) is supercritical for the Cauchy problem (1.1) if and only if p > p c , where p c is defined as p c = m n α , coming from s c = 0 .

Next, we give the local (respectively, global) well-posedness of the solution of the Cauchy problem (1.1) with the initial data in supercritical (respectively, critical) spaces. The similar results have been proved by Weissler [33] for nonlinear heat equations, by Kato [34] for the Navier-Stokes equations, and by Giga and Sawada [35] for the general problems.

Theorem 1.3

(Global and local existence in L p ) (1) (Existence) Let

p ¯ c = max 1 , p c , ( m + 1 ) n n + 1 .

Suppose ( v 0 , w 0 ) L r for a fixed r > p ¯ c or r = p c > 1 . Then, there exist a constant T 0 > 0 and a solution ( V ( t , x ) , W ( t , x ) ) of (1.1) on [ 0 , T 0 ) such that

(1.8) ( t σ V ( t ) , t σ W ( t ) ) B C ( [ 0 , T 0 ) , L p ) ,

with r p < n , 0 σ < 1 m + 1 , or n p < , σ m + n α 1 r 1 n < 1 , where σ = n α 1 r 1 p . And

(1.9) ( t σ W ( t ) L p , t σ W ( t ) L p ) 0 , a s t 0 ,

with r < p < n , 0 < σ < 1 m + 1 or n p < , σ m + n α 1 r 1 n < 1 .

(2) (Estimate for T 0 ) If r > p ¯ c and β ( r ) = m n α r , we have

(1.10) T 0 C ( v 0 , w 0 ) L r m 1 β ( r ) .

(3) (Global existence for small initial data) If there exists a positive constant ε small enough such that ( v 0 , w 0 ) L p c < ε , then T 0 = if p c > 1 and we have

(1.11) ( V ( t ) , W ( t ) ) L p C t σ , 0 < t < ,

with C independent of t, provided that p p c .

(4) (Uniqueness) Solutions of (1.1) satisfying (1.8) and (1.9) for some 0 < σ < 1 m + 1 , p > ( m + 1 ) n n + 1 , and σ = n α 1 r 1 p are unique. If r > p c , σ may equal zero and (1.9) is not necessary to guarantee the uniqueness. In particular if r > p ¯ c , solutions are unique in B C ( [ 0 , T 0 ) , L r ) , provided that r > m + 1 .

We can also prove a slight improvement result if we consider the initial data in critical space L p c ( R n ) .

Corollary 1.1

Assume that α > m and p c > 1 . Let γ ( q ) = n α ( 1 p c 1 q ) . If there exists a constant A such that ( v 0 , w 0 ) L p c ( R n ) with ( v 0 , w 0 ) L p c A , then there is a unique global solution ( V ( t , x ) , W ( t , x ) ) of (1.1) such that for all q [ p c , ) ,

(1.12) ( V ( t , ) , W ( t , ) ) L q C t γ ( q ) ( v 0 L p c + w 0 L p c ) .

Then, using Corollary 1.1, we will consider the case of the initial data with arbitrary high norm in subcritical spaces L p ( R n ) and small norm in critical spaces L p c ( R n ) .

Proposition 1.1

Let ( v 0 , w 0 ) L p c ( R n ) L p ( R n ) with p p c and assume that ( v 0 , w 0 ) L p c A . Let us consider ( V ( t , x ) , W ( t , x ) ) the global solution of (1.1) given by Theorem 1.3. Then,

(1.13) ( V ( t , x ) , W ( t , x ) ) B C ( R + , L p ) B C ( R + , L p c ) ,

(1.14) ( V ( t , ) , W ( t , ) ) L r C t n α ( 1 p 1 r ) ( v 0 L p + w 0 L p ) ,

for all r p and t > 0 .

Using Proposition 1.1, we will prove the following result on the global Cauchy problem with the initial data in supercritical spaces H p s ( R n ) .

Theorem 1.4

Assume that p c > 1 and p ( p c m 1 , p c ] .

  1. There exists a constant A such that, for all ( v 0 , w 0 ) H p s c ( R n ) with ( v 0 , w 0 ) H p s c A , there is a unique global solution ( V ( t , x ) , W ( t , x ) ) of (1.1) in C ( [ 0 , + ) , H p s c ) , which satisfies (1.8)–(1.9). Furthermore, ( V ( t , x ) , W ( t , x ) ) satisfies (1.13) and (1.14).

  2. Let ( v 0 , w 0 ) H p s ( R n ) with s > s c . If ( v 0 , w 0 ) H p s c A , then for the local solution ( V ( t , x ) , W ( t , x ) ) of (1.1) given by Theorem 1.3, we have t ( s s c ) α ( V , W ) B C ( R + , H p s ) and satisfy (1.13) and (1.14).

Remark 1.2

There is no restriction on the size of  ( v 0 , w 0 ) H p s in Part (2) of Theorem 1.4: we just assume that the initial data are small enough in the critical spaces H p s c ( R n ) .

The outline of the rest of the article is as follows. In Section 2, we give several useful lemmas such as the classical Hardy-Littlewood-Sobolev inequality and L p L q and H p s + θ H p s estimates for the semigroup operator e t Λ α . In Section 3, we establish the local existence and uniqueness of mild solution in supercritical spaces H p s ( R n ) , together with the smooth effects and continuous dependence with respect to the initial data. In Section 4, we discuss the local and global Cauchy problem of the mild solution in supercritical spaces L r ( R n ) . In Section 5, we prove the improvement result Corollary 1.1 and consider the global Cauchy problem with the initial data in L p c ( R n ) L p ( R n ) and H p s c ( R n ) H p s ( R n ) , respectively. Finally, we prove Theorem 1.1 in Appendix with tedious calculation.

2 Preliminaries

For the Laplacian operator Δ and the Calderón-Zygmund operator Λ = Δ , we have the following classical Hardy-Littlewood-Sobolev inequality.

Lemma 2.1

[36,37] Let 1 < p < n , and the nonlocal operator ( Δ ) 1 2 is bounded from L p ( R n ) to L n p n p ( R n ) , i.e., f L p ( R n ) ,

( Δ ) 1 2 f L n p n p ( R n ) C ( n , p ) f L p ( R n ) , ( Δ ) 1 f L n p n p ( R n ) C ( n , p ) f L p ( R n ) .

For the fractional power operator Λ α = ( Δ ) α 2 and the semigroup operator e t Λ α , we first consider the Cauchy problem for the homogeneous linear fractional heat equation:

t u + Λ α u = 0 , t > 0 , x R n , u ( x , 0 ) = u 0 ( x ) , x R n .

By the Fourier transform, the solution can be written as

u ( t , x ) = e t Λ α u 0 = 1 ( e t ξ α u 0 ( ξ ) ) = 1 ( e t ξ α ) * u 0 ( x ) = K t ( x ) * u 0 ( x ) ,

where the fractional heat kernel

K t ( x ) = ( 2 π ) n 2 R n e i x ξ e t ξ α d ξ = t n α K ( x t 1 α ) ,

the function K ( x ) L ( R n ) C 0 ( R n ) , where C 0 ( R n ) denotes the space of functions f C ( R n ) satisfying that lim x f ( x ) = 0 .

For the semigroup operator e t Λ α , we have L p L q and H p s + θ H p s estimates [38].

Lemma 2.2

(a) Let 1 p q , α > 0 , and there exists C such that

e t Λ α f L q C ( n , α ) t n α 1 p 1 q f L p , f L p ( R n ) .

(b) Let 1 p q , α > 0 , and γ > 0 . There exists C such that

Λ γ e t Λ α f L q C ( n , α ) t γ α n α 1 p 1 q f L p , f L p ( R n ) .

(c) Let θ 0 , α > 0 , and t ( 0 , T ] . There exists C such that

e t Λ α f H p s + θ C ( T ) t θ α f H p s , f H p s ( R ) .

(d) Let θ 0 , α > 0 , and t ( 0 , T ] . There exists C such that

e t Λ α f H ˙ p s + θ C ( T ) t θ α f H ˙ p s , f H p s ( R ) .

Lemma 2.3

Let p ( 1 , + ) , θ R , and s R . Then,

f H p s 2 C f H p s + θ f H p s θ , f H p s + θ ( R ) .

3 Local Cauchy problem in supercritical space H p s ( R n )

In this section, we prove Theorem 1.2. We first prove the existence of solution in Section 3.1 and then uniqueness in Section 3.2. In Section 3.3, we study the smoothing effects for (1.1), and in Section 3.4, we consider the continuous dependence of the solutions with respect to the initial dada.

3.1 Existence

First, we assume that all conditions in Theorem 1.1 hold true and the initial data ( v 0 , w 0 ) belong to subcritical space H p s ( R n ) . In the sequel, C will denote a positive constant that may be changed from one line to another. To simplify the notations, we have

B ( v , w ) ( t , x ) = 0 t e ( t τ ) Λ α ( v m ( Δ ) 1 ( v w ) ) ( τ ) d τ .

We introduce the exponent p ˜ given by 1 p ˜ = 1 p s n , and by (1.8) and (1.9). Since s > s c , we have p ˜ n ( m + 1 ) n + 1 and p ˜ > p c . We define the spaces Y = C ( [ 0 , T ) , H p s ) and X = C ( [ 0 , T ) , L p ˜ ) . Hence, by the Sobolev embedding theorem, Y X . Now, let us consider the sequence of functions:

(3.1) v 0 = e t Λ α v 0 ( x ) , v j + 1 = v 0 + B ( v j , w j ) , w 0 = e t Λ α w 0 ( x ) , w j + 1 = w 0 + B ( w j , v j ) .

First, we are going to prove that { ( v j , w j ) } converges strongly in X to a limit ( V , W ) , which verifies (1.7) (this proof follows closely Giga’s proof but we detail it for the reader’s convenience), and second, using the new estimates given by Theorem 1.1, we will show that ( V , W ) belongs also to Y . By Part (1) of Lemma 2.2, we have

(3.2) v 0 , w 0 X v 0 , w 0 L p ˜ C v 0 , w 0 H p s .

For ( v , w ) and ( v ˜ , w ˜ ) in X ,

(3.3) B ( v , w ) B ( v ˜ , w ˜ ) = 0 t e ( t τ ) Λ α ( v m ( Δ ) 1 ( v w ) v ˜ m ( Δ ) 1 ( v ˜ w ˜ ) ) ( τ ) d τ = B 1 + B 2 ,

where

B 1 = 0 t e ( t τ ) Λ α ( ( v m v ˜ m ) ( Δ ) 1 ( v w ) ) ( τ ) d τ , B 2 = 0 t e ( t τ ) Λ α ( v ˜ m ( Δ ) 1 ( v v ˜ ( w w ˜ ) ) ) ( τ ) d τ .

Since we are working in the whole Euclidian space R n , the operators e t Λ α and are some Fourier multipliers, and so we have

e t Λ α = e t Λ α = e t Λ α 2 e t Λ α 2 .

Furthermore, by Lemmas 2.2 and 2.3 and Hölder’s inequality, we obtain

B 1 L p ˜ 0 t e ( t τ ) Λ α ( ( v m v ˜ m ) ( Δ ) 1 ( v w ) ) ( τ ) L p ˜ d τ 0 t ( t τ ) β ( v m v ˜ m ) ( Δ ) 1 ( v w ) ( τ ) L q d τ 0 t ( t τ ) β ( v v ˜ ) ( v m 1 + v ˜ m 1 ) ( Δ ) 1 ( v w ) ( τ ) L q d τ 0 t ( t τ ) β v v ˜ L p ˜ ( v L p ˜ m 1 + v ˜ L p ˜ m 1 ) ( Δ ) 1 ( v w ) ( τ ) L n p ˜ n p ˜ d τ 0 t ( t τ ) β v v ˜ L p ˜ ( v L p ˜ m 1 + v ˜ L p ˜ m 1 ) v w L n p ˜ n p ˜ d τ ,

where 1 q = m p ˜ + n p ˜ n p ˜ and by s > s c , β = 1 α + n α 1 q 1 p ˜ = m n α p ˜ < 1 . Then,

(3.4) B 1 X C T 1 β v v ˜ X ( v X m 1 + v ˜ X m 1 ) v w X .

Similarly, we have

(3.5) B 2 X C T 1 β v ˜ X m v v ˜ ( w w ˜ ) X .

Combining (3.3)–(3.5), we have

(3.6) B ( v , w ) B ( v ˜ , w ˜ ) X C T 1 β { v v ˜ X ( v X m 1 + v ˜ X m 1 ) v w X + v ˜ X m v v ˜ ( w w ˜ ) X } .

Due to B ( 0 , 0 ) = 0 and (3.2) and (3.6), we have

v j + 1 X v 0 H p s + C T 1 β v j X m v j w j X , v j + 1 v j X C T 1 β { v j v j 1 X ( v j X m 1 + v j 1 X m 1 ) v j w j X + v j 1 X m v j v j 1 ( w j w j 1 ) X } .

Similarly,

w j + 1 X w 0 H p s + C T 1 β w j X m v j w j X , w j + 1 w j X C T 1 β { w j w j 1 X ( w j X m 1 + w j 1 X m 1 ) v j w j X + w j 1 X m v j v j 1 ( w j w j 1 ) X } .

Then, a standard fixed point argument shows that, for

(3.7) T < C 4 ( v 0 , w 0 ) H p s ν 1 ν = s s c α ,

the sequence { ( v j , w j ) } converges strongly in X to limit ( V , W ) , which obviously solves (1.1) since p ˜ n ( m + 1 ) n + 1 by p ˜ n ( m + 1 ) n + 1 and p ˜ > p c .

Now, we must prove that the limit solution belongs also to Y . Let ( v , w ) Y ; using Theorem 1.1, we have

B ( v , w ) H p s 0 t e ( t τ ) Λ α ( v m ( Δ ) 1 ( v w ) ) ( τ ) H p s d τ C 0 t ( t τ ) 1 α 1 α ( s s m ) v m ( Δ ) 1 ( v w ) ( τ ) H p s d τ C 0 t ( t τ ) β v H p s m v w H p s d τ ,

where β = 1 α + 1 α ( s s m ) = m n α p ˜ , then B ( v , w ) Y C T 1 β v Y m v w Y , and so, by (3.2),

(3.8) v j + 1 Y v 0 H p s + C T 1 β v j Y m v j w j Y .

Similarly, we have

(3.9) w j + 1 Y w 0 H p s + C T 1 β w j Y m v j w j Y .

If T satisfies (3.7), thanks to (3.8)–(3.9), we see that ( v j , w j ) Y remains bounded; thus, we can extract a subsequence ( v j k , w j k ) , which converges weakly-* to a limit ( V ˜ , W ˜ ) Y . Now, the ( v j k , w j k ) converges to ( V , W ) and converges to ( V ˜ , W ˜ ) in Y and so ( V , W ) agrees with ( V ˜ , W ˜ ) . Thus, we have proved the existence of a solution in C ( [ 0 , T ) , H p s ) .

The estimate for T m comes from (3.7), which gives T m C 8 ( v 0 , w 0 ) H p s ν with ν = s s c α .

If T m < + , this explicit lower bound obviously allows us to show the blow-up in H p s norm (one can prove the blow-up in L p ˜ ( R n ) when it holds in H p s ).

3.2 Uniqueness

Let ( V , W ) Y and ( V ˜ , W ˜ ) Y be two solutions for the same initial data ( v 0 , w 0 ) and T < min { T m ( V , W ) , T m ( V ˜ , W ˜ ) } . Then, since ( V , W ) and ( V ˜ , W ˜ ) solve (1.1), respectively, we have V V ˜ X = B ( V , W ) B ( V ˜ , W ˜ ) X . Denote

M = sup t [ 0 , T ] { ( V , W ) L p ˜ , ( V ˜ , W ˜ ) L p ˜ } ,

by (3.6), we obtain

V V ˜ X C T 1 β M m ( V V ˜ X + W W ˜ X ) , W W ˜ X C T 1 β M m ( V V ˜ X + W W ˜ X ) .

So,

V V ˜ X + W W ˜ X C T 1 β M m ( V V ˜ X + W W ˜ X ) .

For T small enough, we have

V V ˜ X + W W ˜ X 1 2 ( V V ˜ X + W W ˜ X ) ,

and so ( V , W ) = ( V ˜ , W ˜ ) on [ 0 , T ] . To conclude, we just have to prove that T m ( V , W ) = T m ( V ˜ , W ˜ ) ; otherwise, we can assume that T m ( V , W ) < T m ( V ˜ , W ˜ ) . By iteration, we can obtain that ( V , W ) = ( V ˜ , W ˜ ) on [ 0 , T m ( V , W ) ] , then

+ = lim t T m ( V , W ) ( V ( t , ) , W ( t , ) ) X = lim t T m ( V , W ) ( V ˜ ( t , ) , W ˜ ( t , ) ) X < ,

which is an apparent contradiction. This ends the proof of uniqueness.

3.3 Smoothing effects

Let ( V , W ) be a solution of (1.1). Using Lemma 2.2, we easily obtain that

V ( t ) e t Λ v 0 H p s + θ C 0 t ( t τ ) 1 α 1 α ( s + θ s m ) V H p s m V W H p s d τ ,

and so, for all θ < α ( 1 β ) = m ( s s c ) , we obtain

V ( t ) e t Λ v 0 H p s + θ C T 1 β θ α V Y m V W Y .

Similarly,

W ( t ) e t Λ w 0 H p s + θ C T 1 β θ α W Y m V W Y .

Then, we prove the smoothing effects.

3.4 Continuous dependence with respect to the initial data

Let ( V , W ) and ( V ˜ , W ˜ ) be two solutions of (1.1) for the initial data ( v 0 , w 0 ) and ( v ˜ 0 , w ˜ 0 ) , respectively. Let

T < min { T m ( V , W ) , T m ( V ˜ , W ˜ ) } , M = sup t [ 0 , T ] { ( V , W ) H p s , ( V ˜ , W ˜ ) H p s } .

Assume that s m 0 , i.e., s m n p ( m + 1 ) 1 m + 1 , one obtains

V V ˜ H p s v 0 v ˜ 0 H p s + C 0 t ( t τ ) β V m ( Δ ) 1 ( V W ) V ˜ m ( Δ ) 1 ( V ˜ W ˜ ) H p s m d τ .

Since s m 0 and q = n p n s m p , we can use the Sobolev embedding L q ( R n ) H p s m ( R n ) , which leads to

V V ˜ H p s v 0 v ˜ 0 H p s + C 0 t ( t τ ) β V m ( Δ ) 1 ( V W ) V ˜ m ( Δ ) 1 ( V ˜ W ˜ ) L q d τ .

By L p L q estimate, we have

V m ( Δ ) 1 ( V W ) V ˜ m ( Δ ) 1 ( V ˜ W ˜ ) L q ( V m V ˜ m ) ( Δ ) 1 ( V W ) L q + V ˜ m ( Δ ) 1 ( V V ˜ ( W W ˜ ) ) L q ( V V ˜ ) ( V m 1 + V ˜ m 1 ) ( Δ ) 1 ( V W ) L q + V ˜ L p ˜ m V V ˜ ( W W ˜ ) L p ˜ V V ˜ L p ˜ ( V L p ˜ m 1 + V ˜ L p ˜ m 1 ) V W L p ˜ + V ˜ L p ˜ m ( V V ˜ L p ˜ + W W ˜ L p ˜ ) V V ˜ H p s ( V H p s m 1 + V ˜ H p s m 1 ) V W H p s + V ˜ H p s m ( V V ˜ H p s + W W ˜ H p s ) .

Thus,

V V ˜ Y v 0 v ˜ 0 H p s + C T 1 β M m ( V V ˜ Y + W W ˜ Y ) .

Similarly,

V V ˜ Y + W W ˜ Y v 0 v ˜ 0 H p s + w 0 w ˜ 0 H p s + 2 C T 1 β M m ( V V ˜ Y + W W ˜ Y ) .

Taking T small enough such that 2 C T 1 β M m 1 2 , we obtain that

(3.10) V V ˜ Y + W W ˜ Y 2 ( v 0 v ˜ 0 H p s + w 0 w ˜ 0 H p s ) .

To conclude, we have to relax our assumption on s . Since ( V , W ) and ( V ˜ , W ˜ ) are the solutions of (1.1),

V V ˜ Y v 0 v ˜ 0 H p s + B ( V , W ) B ( V ˜ , W ˜ ) Y .

By Lemma 2.3, we see that

B ( V , W ) B ( V ˜ , W ˜ ) H p s ( B ( V , W ) H p s + θ + B ( V ˜ , W ˜ ) H p s + θ ) 1 2 B ( V , W ) B ( V ˜ , W ˜ ) H p s θ 1 2 .

One can choose θ < m ( s s c ) such that

(3.11) s c < s θ < m n p ( m + 1 ) 1 m + 1 .

Using the smoothing effects, the first term in the left-hand side of the last inequality is bounded by

{ C T 1 β θ α ( V Y + W Y ) } 1 2 C ( T ) M ( m + 1 ) 2 .

Using (3.10) and (3.11), we bound the second term by

v 0 v ˜ 0 H p s θ 1 2 + V V ˜ H p s θ C ( v 0 v ˜ 0 H p s θ 1 2 + w w 0 H p s θ 1 2 ) C ( v 0 v ˜ 0 Y 1 2 + w w 0 Y 1 2 ) .

Combining these two inequalities, we obtain that

V V ˜ Y C ( T ) ( v 0 v ˜ 0 Y 1 2 + w w 0 Y 1 2 ) .

Thus,

V V ˜ Y + W W ˜ Y C ( T ) ( v 0 v ˜ 0 Y 1 2 + w w 0 Y 1 2 ) .

The proof of Part (3) is complete.

4 Cauchy problem in supercritical space L r ( R n )

In this section, we prove Theorem 1.3, i.e., we describe the local and global Cauchy problem in supercritical spaces L r ( R n ) .

We begin with estimate for e ( t τ ) Λ α ( v m ( Δ ) 1 ( v w ) ) ( τ ) .

(4.1) e ( t τ ) Λ α ( ( v m ( Δ ) 1 ( v w ) ) ( v ˜ m ( Δ ) 1 ( v ˜ w ˜ ) ) ) ( τ ) L s C ( t τ ) 1 α 1 α ( 1 h 1 s ) { ( v m v ˜ m ) ( Δ ) 1 ( v w ) h + v ˜ m ( Δ ) 1 ( v v ˜ ( w w ˜ ) ) L h } C t ( β ( p ) δ ) { v v ˜ L p ( v L p m 1 + v ˜ L p m 1 ) v w L p + v ˜ L p m v v ˜ ( w w ˜ ) L p } ,

where 1 h = m p + 1 p 1 n , β ( p ) = m n α p , and δ = n α 1 s 1 p , ( v , w ) and ( v ˜ , w ˜ ) belong to L p , provided that p > n ( m + 1 ) n + 1 and s h .

Next for T 0 > 0 , we derive a priori estimates for

K j = K j ( T 0 ) = sup 0 < t T 0 t σ ( v j , w j ) L p , j 0 ,

where ( v j , w j ) is defined in (3.1) and σ and p satisfy that

(4.2) σ = n α 1 r 1 p , 0 σ < 1 m + 1 , p > n ( m + 1 ) n + 1 , p r , p p c .

Remark 4.1

It is easy to find a nonempty set of numbers σ , p meeting (4.2). Indeed, the definition of p c shows that m + 1 p c = 1 p c + α n , which gives n α ( 1 r n + 1 n ( m + 1 ) ) < 1 m + 1 for r = p c > 1 . If r > p c , this is obvious. This shows that there exist σ and p satisfying (4.2).

To estimate K j , let us recall the scheme

(4.3) v j + 1 = v 0 + B ( v j , w j ) , w j + 1 = w 0 + B ( w j , v j ) , j N .

Applying (4.1) with ( v , w ) = ( v j , w j ) and ( v ˜ , w ˜ ) = ( 0 , 0 ) to the second term B ( v j , w j ) , we obtain

t σ B ( v j , w j ) L p C t σ 0 t ( t τ ) β ( p ) v j L p m v j w j L p ( τ ) d τ ,

where p > n ( m + 1 ) n + 1 is used. This gives an iterative estimate

(4.4) K j + 1 K 0 + C B K j m + 1 T 0 1 β ( r ) , j N ,

where B = 0 1 ( 1 τ ) β ( p ) τ σ ( m + 1 ) d τ . Since β ( r ) = β ( p ) + σ m , the assumptions σ < 1 m + 1 and p > p c ensure the convergence of B since β ( p c ) = 1 .

For a technical reason, we use a less sharp estimate essentially same as (4.4):

(4.5) K j + 1 K 0 + 2 C B K j m + 1 T 0 1 β ( r ) , j N .

If K 0 or T 0 satisfies that

(4.6) K 0 m T 0 1 β ( r ) 1 2 2 + m C B ,

due to (4.5), an elementary calculation shows that

(4.7) K j < 2 K 0 , j N and 2 C B K j m T 0 1 β ( r ) < 1 2 , j N { 0 } .

We thus have a priori estimate for K j under Condition (4.6).

We next study what conditions on T 0 and ( v 0 , w 0 ) can guarantee (4.6). First, we prove that for σ > 0 ,

(4.8) t σ e t Λ α ( v 0 , w 0 ) L p 0 as t 0 .

Since C c ( R n ) L r ( R n ) is dense in L r , there is a sequence { v 0 i } in C c ( R n ) such that v 0 i v 0 in L r . Then,

t σ e t Λ α v 0 L p t σ e t Λ α ( v 0 v 0 i ) L p + t σ e t Λ α v 0 i L p C v 0 v 0 i L r + t σ e t Λ α v 0 i L p ,

where the constant C is independent of i and t . Since v 0 i L p , we have

e t Λ α v 0 i L p C v 0 i L p .

Thus, t σ e t Λ α v 0 i L p 0 as t 0 since σ > 0 . A similar result is valid for w 0 , and we then have (4.8), which particularly implies that for σ > 0 ,

(4.9) K 0 = sup 0 < t T 0 t σ ( v 0 , w 0 ) L p = sup 0 < t T 0 t σ e t Λ α ( v 0 , w 0 ) L p 0 , as T 0 0 .

If r > p c (consequently β ( r ) < 1 ), then Condition (1.10) ensures (4.6) since K 0 ( v 0 , w 0 ) L r is bounded independent of ( v 0 , w 0 ) and T 0 . In the case r = p c , (4.9) shows that for small T 0 , we have (4.6) for every T 0 > 0 . Moreover, if β ( p c ) = 1 , (4.6) does not contain T 0 explicitly, so K j is bounded on ( 0 , T ) even if T = . Therefore, (4.6) holds under the assumptions on ( v 0 , w 0 ) and T 0 in (1)–(3) of Theorem 1.3.

To see the existence, it remains to prove the convergence of { ( v j , w j ) } as j . Actually, we shall first prove that t σ ( v j , w j ) converges in B C ( [ 0 , T 0 ) , L p ) provided that p and σ satisfy (4.2). Note that (4.7) implies that t σ ( v j , w j ) B C ( [ 0 , T 0 ) , L p ) . Moreover, from (4.6), (4.7), and (4.9), it follows that if σ > 0 , t σ ( v j , w j ) tends to zero as t 0 . To show the convergence, we consider the successive difference of ( v j , w j ) constructed by (4.3):

v j + 1 v j = B ( v j , w j ) B ( v j 1 , w j 1 ) , w j + 1 w j = B ( w j , v j ) B ( w j 1 , v j 1 ) .

Just like deriving (4.4), applying (4.1) with s = p , we have

t σ v j + 1 v j L p + t σ w j + 1 w j L p 2 C B T 0 1 β ( r ) K j m { sup 0 < τ T 0 τ σ v j v j 1 L p + sup 0 < τ T 0 τ σ w j w j 1 L p } ,

where p > p c is used. Since 2 C B T 0 1 β ( r ) K j m < 1 2 < 1 , this shows that there is a pair of functions ( V , W ) such that lim j t σ ( v j , w j ) = t σ ( V , W ) in B C ( [ 0 , T 0 ) , L p ) , which solves

V = v 0 + B ( V , W ) and W = w 0 + B ( W , V ) .

Also, if σ > 0 , t σ ( V , W ) L p tends to zero as t 0 since each t σ ( v j , w j ) have the same property.

To complete the proof of (1.8) and (1.9), we have to relax the condition on p . Let p [ r , p ] and σ = n α ( 1 r 1 p ) . We shall prove that t σ ( v j , w j ) converges in B C ( [ 0 , T 0 ) , L p ) and that t σ ( V , W ) L p tends to zero as t 0 if σ > 0 . Applying (4.1) with s = p yields

(4.10) t σ v j + 1 v j L p + t σ w j + 1 w j L p 2 C T 0 1 β ( r ) K j m { sup 0 < τ T 0 τ σ v j v j 1 L p + sup 0 < τ T 0 τ σ w j w j 1 L p } ,

with a different constant C . Since t σ ( v j , w j ) converges in B C ( [ 0 , T 0 ) , L p ) , this implies

t σ ( v j , w j ) t σ ( V , W ) in B C ( [ 0 , T 0 ) , L p ) .

Thus, we have proved (1.8). Since t σ ( v j , w j ) L p and t σ ( v 0 , w 0 ) L p both tend to zero as t 0 by (4.8), (4.10) implies (1.9). The asymptotic behavior (1.11) comes from (4.5) if p and σ satisfy (4.2). For general p in Theorem 1.3, it is not difficult to see that (1.11) also holds by using (4.8) and (4.10). Thus, we have proved (1)–(3) of Theorem 1.3.

It remains to prove the uniqueness (4). Let ( V , W ) and ( V ˜ , W ˜ ) be two solutions of (1.1) satisfying assumptions of (4). We may assume that ( V , W ) and ( V ˜ , W ˜ ) satisfy (1.8)–(1.9) for same σ and p such that 0 < σ < 1 m + 1 , p > n ( m + 1 ) n + 1 , and σ = n α 1 r 1 p . If r > p c , σ may equal zero and we assume (1.8) only. In fact, if ( V , W ) also satisfy (1.8)–(1.9) for every σ and p such that 0 σ < σ , σ = n α ( 1 r 1 p ) . To see this, we just use the estimate

t σ V L p t σ v 0 L p + C T 0 1 β ( r ) ( sup 0 < τ T 0 τ σ V L p ) m ( sup 0 < τ T 0 τ σ V W L p ) ,

which is proved similar to (4.10).

We first consider the case r > p c . Let K be a constant such that

t σ ( V , W ) L p , t σ ( V ˜ , W ˜ ) L p K , 0 t T 0 ,

where σ and p satisfy 0 σ < 1 m + 1 , p > n ( m + 1 ) n + 1 , and σ = n α 1 r 1 p . Since

V V ˜ = B ( V , W ) B ( V ˜ , W ˜ ) , W W ˜ = B ( W , V ) B ( W ˜ , V ˜ ) ,

by definition, we have the estimate for 0 τ t 0 ,

(4.11) t σ V V ˜ L p + t σ W W ˜ L p 2 C t 0 1 β ( r ) K m sup 0 τ t 0 τ σ ( V V ˜ L p + W W ˜ L p ) ,

which follows from (4.1) similar to (4.10). Since r > p c , we can take t 0 small so that 2 C t 0 1 β ( r ) K m < 1 . This implies that V = V ˜ on [ 0 , t 0 ) . Since ( V , W ) , ( V ˜ , W ˜ ) B C ( [ t 0 , T 0 ) , L p ) . The aforementioned argument with initial data V ( t 0 ) = V ˜ ( t 0 ) and W ( t 0 ) = W ˜ ( t 0 ) shows that ( V , W ) = ( V ˜ , W ˜ ) on [ t 0 , t 0 + t 0 ) . By iterating, we have ( V , W ) = ( V ˜ , W ˜ ) on [ 0 , T 0 ) .

It remains to discuss the case r = p c . Let K ( t 0 ) be a constant such that

t σ ( V , W ) L p , t σ ( V ˜ , W ˜ ) L p K ( t 0 ) , 0 t t 0 ,

where 0 σ < 1 m + 1 , p > n ( m + 1 ) n + 1 , and σ = n α 1 r 1 p . Here, by (1.9), K ( t 0 ) tends to zero as t 0 0 . Instead of (4.11), for 0 τ t 0 , we have

t σ V V ˜ L p + t σ W W ˜ L p 2 C K ( t 0 ) m sup 0 τ t 0 τ σ ( V V ˜ L p + W W ˜ L p ) ,

since β ( p c ) = 1 . Take t 0 > 0 small so that 2 C K ( t 0 ) m < 1 . As is seen in the preceding paragraph, we have ( V , W ) = ( V ˜ , W ˜ ) on [ 0 , t 0 ) . By iterating as earlier, we have ( V , W ) = ( V ˜ , W ˜ ) on [ 0 , T 0 ) . This concludes the proof of uniqueness.

Finally, we prove that for p > n , the mild solution of the Cauchy problem (1.1) satisfies the estimates (1.8)–(1.9), too. Of course, we first derive a priori estimate for

K j = K j ( T 0 ) = sup 0 < t T 0 t σ ( v j , w j ) L p , j 0 , T 0 > 0 ,

the argument is similar as earlier, here σ and p satisfy that

σ = n α 1 r 1 p , σ m + n α 1 r 1 n < 1 , m p , n p .

Thus, we obtain the existence of the solution. Then, we relax the index p m as earlier. Combining with the uniqueness of solution, we finish the proof of Theorem 1.3.

5 Global Cauchy problem in subcritical space L q ( R n )

In this section, we prove Theorem 1.4. We study the global Cauchy problem for small initial data in L p c ( R n ) . First, in Section 5.1, we study the case of initial data that belong only to L p c ( R n ) and we prove Corollary 1.1. In Section 5.2, we study the global Cauchy problem for initial data in L p ( R n ) L p c ( R n ) when L p ( R n ) is subcritical for (1.1) and we prove Proposition 1.1. Finally, in Section 5.3, we consider initial data in H p s ( R n ) space, and then, we prove Theorem 1.4.

5.1 Initial data in L p c ( R n )

Let us consider ( v 0 , w 0 ) L p c ( R n ) . In Theorem 1.3, we have proved that there exists a non-negative absolute constant A such that if ( v 0 , w 0 ) L p c A , then there exists a unique global solution ( V , W ) B C ( R + , L q ) to Problem (1.1), which satisfies

(5.1) t t γ ( q ) ( V ( t , ) , W ( t , ) ) L q B C ( R + ) ,

for all q and γ ( q ) such that p c q < n and 0 γ ( q ) < 1 m + 1 , and

(5.2) lim t 0 + t γ ( q ) ( V ( t , ) , W ( t , ) ) L q = 0 ,

for all q and γ ( q ) such that

(5.3) p c < q < n , q > ( m + 1 ) n n + 1 and 0 < γ ( q ) < 1 m + 1 .

First, we are going to prove that for p c q < n and 0 < γ ( q ) < 1 m + 1 ,

(5.4) ( V ( t , ) , W ( t , ) ) L q C t γ ( q ) ( v 0 L p c + w 0 L p c ) ,

which is a little more precise than the estimate ( V , W ) L q C t γ ( q ) .

Second, we are going to relax the restriction γ ( q ) < 1 m + 1 in this estimate. Indeed, when α < m ( m + 1 ) , it is easy to check that γ ( q ) < 1 m + 1 is fulfilled for all q [ p c , n ) and so, the asymptotic estimate (5.4) too. On the contrary, when α m ( m + 1 ) , one must assume that q p c , 1 p c α n ( m + 1 ) 1 to be sure that γ ( q ) < 1 m + 1 holds. So, when α < m ( m + 1 ) , the asymptotic estimates are proved only for q in the range p c , 1 p c α n ( m + 1 ) 1 and we want to show that they hold for all exponents q [ p c , n ) .

To prove Corollary 1.1, let us come back to the proof of Theorem 1.3. In the critical case (when ( v 0 , w 0 ) L p c ( R n ) ), to prove the existence of solutions for (1.1), one introduce, for p c < q < n , q > ( m + 1 ) n n + 1 and 0 < γ ( q ) < 1 m + 1 , the Banach spaces

X q = { f ( t , x ) : t t γ ( q ) f ( t , x ) L q B C ( R + ) } , Y = { f ( t , x ) : t f ( t , x ) L p c B C ( R + ) } .

Then, for the sequence { ( v j , w j ) } defined in (3.1), we have the estimates

(5.5) v j + 1 X q C 1 v 0 X q + C 2 v j X q m v j w j X q , w j + 1 X q C 1 w 0 X q + C 2 w j X q m v j w j X q ,

where f ( t , x ) X q = sup t > 0 f ( t , x ) L q . Then, when ( v 0 , w 0 ) L p c A , using (5.5) and (5.2)–(5.3), one can prove that the { ( v j , w j ) } converge in X q to ( V ( t , x ) , W ( t , x ) ) the unique solution of (1.1) such that (5.2) and (5.3) are fulfilled. Furthermore, to prove that ( V ( t , x ) , W ( t , x ) ) also belong to B C ( R + , L p c ) , one can easily check that

B ( V , W ) Y C V X q m V W X q

as soon as p c < q < n , q > ( m + 1 ) n n + 1 , and 0 < γ ( q ) < 1 m + 1 .

Now, let us come back to the proof of Corollary 1.1. By (5.5), it is obvious that the sequence { ( v j , w j ) } stay in the ball B ( 0 , 2 R ) for the X q topology as soon as C 2 ( 2 C 1 R ) m + 1 C 1 R with R = max { v 0 X q , w 0 X q } , which holds for

max { v 0 X q , w 0 X q } 1 2 C 1 1 2 C 2 1 m .

Now, by Lemma 2.2, we have

v 0 X q C v 0 L p c , w 0 X q C w 0 L p c ,

and for

r max { v 0 L p c , w 0 L p c } 1 2 C C 1 1 2 C 2 1 m ,

there exists a global solution ( v ( t , x ) , w ( t , x ) ) of (1.1), which belongs to the ball

B ( 0 , 2 C 1 R ) B ( 0 , 2 C C 1 r ) ,

for the X q topology. Thus, the proof of Corollary 1.1 is completed for the exponent q such that p c < q < n , q > ( m + 1 ) n n + 1 , and 0 < γ ( q ) < 1 m + 1 . If we can deal with the special case of L p c ( R n ) norm, we finish the proof completely.

Now, we are going to prove that the asymptotic estimates

( V ( t , ) , W ( t , ) ) L q C t γ ( q ) ( v 0 L p c + w 0 L p c )

hold also when γ ( q ) 1 m + 1 . First, for ( v 0 , w 0 ) such that ( v 0 , w 0 ) L p c A , let us consider ( v ( t , x ) , w ( t , x ) ) the solution of (1.1). Let us consider q 0 an exponent such that q 0 > p c and γ ( q 0 ) 1 m + 1 (such a q 0 always exists since p c > 1 : see the remark after Theorem 1.3). Next, let us consider the sequence q i defined by

(5.6) n α 1 q i 1 q i + 1 = δ i < 1 m + 1 ,

and note that { q i } is increasing and there exists q k such that n α q k < 1 m + 1 .

Let us define

I ( q i , q i + 1 ) = 0 1 ( 1 τ ) m n α q i + 1 τ ( m + 1 ) δ i d τ .

Then, by (5.6), for all i 0 , I ( q i , q i + 1 ) < + . Now, we pick t 0 > 0 and we consider ( V , W ) the solution of

(5.7) V ( t , x ) = e t Λ α V 0 + B ( V , W ) , W ( t , x ) = e t Λ α W 0 + B ( W , V ) , V ( 0 , x ) = V 0 ( x ) = V ( t 0 , x ) , W ( 0 , x ) = W 0 ( x ) = W ( t 0 , x ) .

Similar to the previous steps and γ ( q 0 ) 1 m + 1 , we have ( V 0 , W 0 ) L p c ( R n ) L q 0 ( R n ) with

(5.8) ( V 0 , W 0 ) L q 0 C t γ ( q 0 ) ( V 0 L p c + W 0 L p c ) .

Lemma 5.1

Let T = T ( t 0 ) such that

(5.9) ( 2 C ) m + 1 T 1 p c q 0 I ( q 0 , q 1 ) ( V 0 L q 0 + W 0 L q 0 ) < 1 ,

then for all t [ 0 , T ) , we have

(5.10) ( V ( t , ) , W ( t , ) ) L q 1 C t δ 0 ( V 0 L q 0 + W 0 L q 0 ) .

Proof

Indeed, since ( V 0 ( x ) , W 0 ( x ) ) L q 0 ( R n ) with q 0 > p c , following the proof of Theorem 1.3, we see that the sequence

v 0 = e t Λ α V 0 , v j + 1 = v 0 + B ( v j , w j ) , w 0 = e t Λ α W 0 , w j + 1 = w 0 + B ( w j , v j ) .

converges strongly to ( V ˙ , W ˙ ) in C ( [ 0 , T ) , L q 0 ( R n ) ) . By Lemma 2.2, ( v 0 , w 0 ) obviously satisfies (5.10) for all t > 0 . Now, if v j satisfies (5.10), then

v j + 1 ( t ) L q 1 C V 0 L q 0 + C 0 t ( t τ ) m n α q 1 v j L q 1 m v j w j L q 1 d τ C V 0 L q 0 + C 2 m + 1 C m + 1 ( V 0 L q 0 + W 0 L q 0 ) m + 1 0 t ( t τ ) m n α q 1 τ ( m + 1 ) δ 0 d τ

for all t ( 0 , T ) , and so

v j + 1 ( t ) L q 1 2 C t δ 0 ( V 0 L q 0 + W 0 L q 0 ) 1 2 + 2 m C m + 1 ( V 0 L q 0 + W 0 L q 0 ) m I ( q 0 , q 1 ) T 1 p c q 0 , t ( 0 , T ) .

Hence, if T satisfies (5.9),

( v j + 1 ( t , ) , w j + 1 ( t , ) ) L q 1 2 C t δ 0 ( V 0 L q 0 + W 0 L q 0 ) .

So, by introduction, (5.10) holds for all j N , and thus, Lemma 5.1 is proved.□

Using the uniqueness result in the supercritical case and (5.7), we see that

V ( t , x ) = V ( t + t 0 , x ) , W ( t , x ) = W ( t + t 0 , x ) ,

due to Lemma 5.1, for each t [ 0 , T ( t 0 ) ) and T ( t 0 ) satisfies (5.9), we have

( V ( t + t 0 , ) , W ( t + t 0 , ) ) L q 1 2 C t δ ( V ( t 0 ) L q 0 + W ( t 0 ) L q 0 ) .

Now, we claim that there exists an absolute constant A such that, when ( v 0 , w 0 ) L p c A , one can always take T ( t 0 ) = 1 2 in the previous inequality. Indeed, by Lemma 5.1, we have only to make sure that

( 2 C ) m + 1 t 0 2 1 p c q 0 I ( q 0 , q 1 ) ( V 0 L q 0 + W 0 L q 0 ) m < 1 ,

combining with (5.8) and following by

( 2 C ) m + 1 t 0 2 1 p c q 0 m γ ( q 0 ) I ( q 0 , q 1 ) ( v 0 L p c + w 0 L p c ) m < 1 ,

and 1 p c q 0 m γ ( q 0 ) = 0 , it is sufficient to make sure that

( 2 C ) m + 1 I ( q 0 , q 1 ) ( v 0 L p c + w 0 L p c ) m < 1 .

Thus, when ( v 0 , w 0 ) is small enough in L p c ( R n ) , (5.9) holds for each t 0 > 0 , and so

V 3 t 0 2 , W 3 t 0 2 L q 1 4 C t 0 δ 0 t 0 γ ( q 0 ) ( v 0 L p c + w 0 L p c ) ,

and since t 0 is arbitrary, then for all t > 0 , we have

( V ( t ) , W ( t ) ) L q 1 4 C t n α 1 p c 1 q 1 ( v 0 L p c + w 0 L p c ) .

Now, since I ( q i , q i + 1 ) < and the required estimate for q 1 defined by (5.6) is proved, we have just to iterate this proof to obtain the required estimate in L q 2 ( R n ) -norm. Thus, for each q i , the proof follows by induction. Now, if q ( q i , q i + 1 ) , we obtain the result by interpolation. Thus, we have proved that the global solution ( V ( t , x ) , W ( t , x ) ) of (1.1) satisfies that

( V ( t ) , W ( t ) ) L q 1 C t γ ( q ) ( v 0 L p c + w 0 L p c ) , q [ p c , n ) .

For the case p [ n , + ) , the proof is similar as earlier, we omit the detail.

5.2 Initial data in L p c ( R n ) L p ( R n )

Let p < p c . For the initial data ( v 0 , w 0 ) L p c ( R n ) L p ( R n ) such that ( v 0 , w 0 ) L p c A , ( V ( t , x ) , W ( t , x ) ) is the mild solution of (1.1), which belongs to B C ( R + , L p c ) and satisfies Estimates (5.1)–(5.3). Using the slight improvement about the decay of the L q ( R n ) norms (Estimates (1.12) of Corollary 1.1) that we previously proved, we are first going to show that the solution belongs to L p ( R n ) for all t (Step 1), then we will prove that ( V , W ) belongs to B C ( R n , L p ( R n ) ) (Step 2) and next, ( V , W ) satisfies the asymptotic estimates (1.12).

Step 1. For ( v 0 , w 0 ) L p c ( R n ) L p ( R n ) and we want to prove that, for any T > 0 ,

(5.11) ( V ( t ) , W ( t ) ) L p C ( T ) , t [ 0 , T ] .

First, assume that

(5.12) max 1 , p c m p < p c .

Since ( V ( t ) , W ( t ) ) is a solution for (1.1), for all T > 0 and t [ 0 , T ] ,

V ( t ) L p C v 0 L p + B ( V , W ) L p C v 0 L p + C 0 t ( t τ ) 1 α n α 1 r 1 p V m ( Δ ) 1 ( V W ) L r d τ C v 0 L p + C 0 t ( t τ ) n α q V L p m m V W L q d τ ,

where 1 r = m p m + 1 q 1 n and p c q < n . If we choose q such that q n , by (5.12), p m > p c , using Estimate (1.12) of Corollary 1.1, we obtain that

V ( t ) L p C v 0 L p + C 0 t ( t τ ) n α q τ m γ ( p m ) γ ( q ) ( v 0 L p c + w 0 L p c ) m + 1 d τ , C v 0 L p + C ( T ) ( v 0 L p c + w 0 L p c ) m + 1 ,

where p c q < n , for all p c m p < p c , we have 0 < m γ ( p m ) + γ ( q ) < 1 n α q < 1 .

Similarly,

W ( t ) L p C v 0 L p + C ( T ) ( v 0 L p c + w 0 L p c ) m + 1 .

Thus, we obtain Estimate (1.14) for p satisfying (5.12), and if p c < m , the proof is complete.

Assume now that

(5.13) max 1 , p c m 2 p < p c m .

First, if ( v 0 , w 0 ) L p c ( R n ) L p ( R n ) , then ( v 0 , w 0 ) L q ( R n ) for all [ p c m , p c ) , and then, by the previous result, ( v ( t ) , w ( t ) ) L q C ( T , v 0 , w 0 ) for all q in the range [ p c m , p c ) .

Second, since ( v ( t , x ) , w ( t , x ) ) is a solution of (1.1),

V ( t ) L p C v 0 L p + B ( V , W ) L p C v 0 L p + C 0 t ( t τ ) 1 α n α 1 r 1 p V m ( Δ ) 1 ( V W ) L r d τ C v 0 L p + C 0 t ( t τ ) n α q V L p m m V W L q d τ ,

where 1 r = m p m + 1 q 1 n and p c q < n . Choosing q such that q n , by (5.13), p m [ p c m , p c ) . Hence, by the previous result, we can use the bound ( V ( t , x ) , W ( t , x ) ) L p m C ( T , v 0 , w 0 ) , which leads to

V ( t ) , W ( t ) L p C v 0 , w 0 L p + C ( T , v 0 , w 0 ) .

Thus, Estimate (5.11) holds for all p in the range [ max { 1 , p c m 2 } , p c ) .

Finally, for p [ p c m n 1 , p c m n ) , the proof of (5.11) follows easily by induction.

Step 2. In Step 1, we have proved that ( V ( t , x ) , W ( t , x ) ) is the solution of (1.1) that belongs to L p ( R n ) for all t 0 when ( v 0 , w 0 ) belongs to L p ( R n ) L p c ( R n ) and when ( v 0 , w 0 ) is small enough in L p c ( R n ) . Now, we are going to prove that ( V ( t , x ) , W ( t , x ) ) belongs to B C ( R + , L p ( R n ) ) . Let us consider T > 0 and t in [ 0 , T ] . First, since ( V , W ) is a mild solution of (1.1), by Lemma 2.2,

V ( t ) L p C v 0 L p + B ( V , W ) L p C v 0 L p + C 0 t ( t τ ) ξ ( q ) V m ( Δ ) 1 ( V W ) L q d τ ,

where q is any exponent in [ 1 , p ) which will be fixed later, and where ξ ( q ) is defined by

(5.14) ξ ( q ) = 1 α + n α ( 1 q 1 p ) .

Using Hölder’s inequality, we obtain

V ( t ) L p C v 0 L p + C 0 t ( t τ ) ξ ( q ) V L q q 1 V L q q 2 ( m 1 ) m 1 V W L q 3 d τ ,

where 1 q = 1 q q 1 + 1 q q 2 ( m + 1 ) + 1 q 3 1 n and p c q 3 < n . Furthermore, we choose q 1 such that q q 1 = p to obtain

V ( t ) L p C v 0 L p + C V L ( [ 0 , ) , L p ) 0 t ( t τ ) ξ ( q ) V L q q 2 ( m 1 ) m 1 V W L q 3 d τ .

Now, if we choose q such that q p with q < p and choose q 3 such that q 3 = n with q 3 < n then, since q q 1 = p , i.e., q 1 1 and q 3 is large enough. Hence, it follows that q q 2 ( m + 1 ) p c . Next, for q q 2 ( m + 1 ) p c , by Corollary 1.1 and the L q q 2 ( m 1 ) ( R n ) -norm, we have

V ( t ) L q q 2 ( m 1 ) C t γ ( q q 2 ( m 1 ) ) ( v 0 L p c + w 0 L p c ) ,

and by L q 3 ( R n ) -norm, we obtain

V W L q 3 C t γ ( q 3 ) ( v 0 L p c + w 0 L p c ) .

Thus,

V ( t ) L p C v 0 L p + C V L ( [ 0 , T ] , L p ) ( v 0 L p c + w 0 L p c ) m 0 t ( t τ ) ξ ( q ) τ θ ( q ) d τ ,

where

(5.15) θ ( q ) = ( m 1 ) γ ( q q 2 ( m 1 ) ) + γ ( q 3 ) = 1 n α 1 q q 2 + 1 q 3 1 1 α ,

since q 3 n and q 2 is large enough, when we choose q p . One can easily check that 0 < ξ ( q ) < 1 and 0 < θ ( q ) < 1 (since α > m 1 ). Furthermore, ξ ( q ) + θ ( q ) = 1 , and so,

(5.16) V ( t ) L p C v 0 L p + C V L ( [ 0 , T ] , L p ) ( v 0 L p c + w 0 L p c ) m .

Now, if ( v 0 , w 0 ) L p c is small enough, then C ( v 0 L p c + w 0 L p c ) m 1 2 , then, by (5.16) and V L ( [ 0 , T ] , L p ) < for all T > 0 , we have

V L ( [ 0 , T ] , L p ) v 0 L p 1 C ( v 0 L p c + w 0 L p c ) m 2 v 0 L p .

To conclude, we have just to remark that the right-hand side of this estimate does not depend of T . Thus, we have proved that ( V , W ) , the mild solution of (1.1) that belongs to B C ( R + , L p ( R n ) ) .

Step 3. Now, we have to prove the L r ( R n ) Estimate (1.14) of Proposition 1.1. They hold obviously for the term ( e t Λ α v 0 , e t Λ α w 0 ) by Lemma 2.2; hence, we just deal with the nonlinear term B ( V , W ) and B ( W , V ) . First, let us suppose that

(5.17) δ ( r ) = n α 1 p 1 r < 1 1 α .

Then,

B ( V , W ) L r C 0 t ( t τ ) δ ( r ) e 1 2 ( t τ ) Λ α ( V m ( Δ ) 1 ( V W ) ) L p d τ C 0 t ( t τ ) δ ( r ) ξ ( q ) V L q q 1 V L q q 2 ( m 1 ) m 1 V W L q 3 d τ ,

where q [ 1 , p ) , p c q 3 < n , 1 q = 1 q q 1 + 1 q q 2 ( m + 1 ) + 1 q 3 1 n , and ξ ( q ) is given by (5.14). Now, taking q q 1 = p with q p , then q 1 1 , taking q 3 = n with p c q 3 < n and q q 2 ( m + 1 ) p c , and so using Corollary 1.1, we obtain

B ( V , W ) L r C ( sup t R + V L p ) ( v 0 L p c + w 0 L p c ) m 0 t ( t τ ) δ ( r ) ξ ( q ) τ θ ( q ) d τ ,

where θ ( q ) is given by (5.15). If (5.17) holds, then one can choose q , q 1 , and q 2 such that 0 < θ ( q ) < 1 , 0 < δ ( r ) + ξ ( q ) < 1 , and ξ ( q ) + θ ( q ) = 1 , and so

B ( V , W ) L r C t δ ( r ) ( sup t R + V L p ) ( v 0 L p c + w 0 L p c ) m .

Then, since v L p C v 0 L p (by Step 2), we have B ( V , W ) L r C t δ ( r ) v 0 L p .

Now, if (5.17) is not fulfilled, we build a sequence { r i } defined by

r 0 = p , n α 1 r i 1 r i + 1 = δ i < max 1 1 α , m + 1 .

Now, if p < r 1 < r 2 < p c , since ( V , W ) is bounded in L p L p c , then ( V , W ) is also bounded in L r for all r in [ p , p c ] and for each t 0 .

Let t 0 > 0 and ( V , W ) be the solution of (5.7). We have already proved that

( V 0 , W 0 ) L p c ( R n ) L r 1 ( R n ) ,

with V 0 L r 1 C t 0 δ ( r 1 ) v 0 L p , W 0 L r 1 C t 0 δ ( r 1 ) w 0 L p , and furthermore, ( V ( t , ) , W ( t , ) ) is bounded in L p c ( R n ) L r 1 ( R n ) . So we just have to iterate the previous proof to estimate V ( t 0 , x ) = V ( 2 t 0 , x ) and W ( t 0 , x ) = W ( 2 t 0 , x ) in L r 2 ( R n ) norm with respect to V 0 ( x ) = V ( t 0 , x ) and W 0 ( x ) = W ( t 0 , x ) in L r 1 ( R n ) norm to obtain the required estimate and we can do this until r i p c .

Now, let us denote by N the first index such that r N = p c . We have proved that

(5.18) V ( t ) L N r C t δ ( r N ) v 0 L p and W ( t ) L N r C t δ ( r N ) w 0 L p ,

where ( V , W ) is the mild solution of (5.7) with the initial data replaced by

V 0 ( x ) = V ( ( N 1 ) t 0 , x ) and W 0 ( x ) = W ( ( N 1 ) t 0 , x ) .

Then, we can choose r N + 1 such that p c < r N + 1 < n . Followed by Corollary 1.1, we have

(5.19) V ( t ) L r N + 1 C t γ ( r N + 1 ) ( V 0 L p c + W 0 L p c ) , W ( t ) L r N + 1 C t γ ( r N + 1 ) ( V 0 L p c + W 0 L p c ) ,

where γ ( r N + 1 ) is defined in Corollary 1.1 and ( V ( t ) , W ( t ) ) is the mild solution of (5.7) with the initial data replaced by V 0 ( x ) = V ( N t 0 , x ) and W 0 ( x ) = W ( N t 0 , x ) . Then, combining (5.18)–(5.19), we have

V ( t 0 ( N + 1 ) ) L r N + 1 C t 0 δ ( r N + 1 ) ( v 0 L p + w 0 L p ) , W ( t 0 ( N + 1 ) ) L r N + 1 C t 0 δ ( r N + 1 ) ( v 0 L p + w 0 L p ) .

This ends the proof of Proposition 1.1 since t 0 > 0 is arbitrary.

5.3 Initial data in H p s ( R n )

Let us consider an initial data ( v 0 , w 0 ) such that ( v 0 , w 0 ) H p s c A . Then, by the Sobolev embedding theorem, ( v 0 , w 0 ) L p c ( R n ) L p ( R n ) and if A is small enough, then ( v 0 , w 0 ) L p c A . So, according to Proposition 1.1, there exists a unique global solution ( V ( t , x ) , W ( t , x ) ) of (1.1), and this solution satisfies (1.13) and (1.14). Hence, to prove that ( V , W ) belongs to B C ( R + , H p s c ) , we have only to check that ( V , W ) remains bounded in the homogeneous space H ˙ p s c ( R n ) , thanks to the following well-known inequality:

f H p s C ( f L p + f H ˙ p s ) , s 0 .

Now, since ( V , W ) is a solution of (1.1),

V ( t ) H ˙ p s c C v 0 H p s c + B ( V , W ) H ˙ p s c C v 0 H p s c + C 0 t ( t τ ) 1 α s c α e 1 2 ( t τ ) Λ α V m ( Δ ) 1 ( V W ) L p d τ C v 0 H p s c + C 0 t ( t τ ) 1 + s c α n α 1 r 1 p V m ( Δ ) 1 ( V W ) L r d τ C v 0 H p s c + C 0 t ( t τ ) λ ( q ) V L q m V W L q 1 d τ ,

where

λ ( q ) = 1 + s c α + n α m q + 1 q 1 1 n 1 p ,

with q ( p c , p m ] and q 1 [ p c , n ) .

Since p > p c m 1 , one can easily check that s c < α 1 1 m < α 1 ( m < α ). Taking q 1 n with q 1 < n and q p m , one can always choose q such that 0 < λ ( q ) < 1 . Then, for this choice of q , we obtain

V ( t ) H ˙ p s c C v 0 H p s c + C ( sup t R + t γ ( q ) V L q ) m ( sup t R + t γ ( q 1 ) V W L q 1 ) 0 t ( t τ ) λ ( q ) τ m γ ( q ) γ ( q 1 ) d τ C v 0 H p s c + C ( sup t R + t γ ( q ) V L q ) m ( sup t R + t γ ( q 1 ) V W L q 1 ) ,

since γ ( q ) + m γ ( q ) + γ ( q 1 ) = 1 and m γ ( q ) + γ ( q 1 ) ( 0 , 1 ) . But by Corollary 1.1, we know that t γ ( q ) V L q remains bounded for all t > 0 and so ( V , W ) belongs to B C ( R n , H ˙ p s c ) . Thus, we have proved that ( V , W ) belongs to B C ( R + , H p s c ) . This ends the proof of Part (1).

Now, let ( v 0 , w 0 ) H p s ( R ) such that ( v 0 , w 0 ) H p s c A . Then, according to Part (1) of Theorem 1.2 and to Part (1) of Theorem 1.4, there exists a unique solution of (1.1) in C ( [ 0 , T m ) , H p s ) B C ( R + , H p s c ) and so, to prove Part (2), we must show that blow-up in H p s ( R n ) norm cannot occur when t T m . But, like the proof of Part (1), one can easily show that for T > 0 , t [ T m , T m + T ] ,

(5.20) V ( t ) e t Λ α v 0 H p s c + θ C 0 t ( t τ ) η ( θ ) V p c m V W L q d τ ,

where η ( θ ) = 1 + s c + θ α + n α ( m p c ) , 0 < θ < α m 1 , and q n with q < n . By H p s c ( R n ) L p c ( R n ) and (5.4), we obtain that the right-hand side of (5.20) is bounded by

C 0 t ( t τ ) η ( θ ) V H p s c m τ γ ( q ) ( v 0 L p c + w 0 L p c ) d τ C ( sup t R n V H p s c m ) m 0 t ( t τ ) η ( θ ) τ γ ( q ) d τ C t θ α ( sup t R n V H p s c m ) m ,

since 0 < η ( θ ) < 1 , 0 < γ ( q ) < 1 , and η ( θ ) + γ ( q ) = 1 + θ α , where γ ( q ) is given by Corollary 1.1. Hence, if blow-up holds in H p s c + θ ( R n ) , it holds in H p s c ( R n ) norm, this contradicts Part (1). Now since s > s c is arbitrary, we have just to iterate like the proof of Part (2) of Theorem 1.2 with θ < m ( s s c ) replaced by θ = m ( s s c ) .

6 Proof of Theorem 1.1

In this section, we prove Theorem 1.1, the main part is the following nonlinear estimate:

F ( v , w ) H p s m C v H p s m v w H p s

that we used in a crucial way in the proof of Theorem 1.2 (our result about local existence and uniqueness for the Cauchy problem (1.1)). First, we are going to consider the case (H1) (i.e., when s m 0 ), Then, after recalling a few results about Littlewood-Paley analysis, we will prove Theorem 1.1 when the assumption (H2) is fulfilled ( 0 < s m < m ).

6.1 Case s m 0

Here, we suppose that max 0 , n p n + 1 m + 1 < s and that (H1) is fulfilled, i.e. s m < 0 . Now, consider ( v , w ) H p s ( R n ) . Then, by the Sobolev embedding theorem ( s m 0 and p ( 1 , + ) ), we have

(6.1) L q ( R n ) H p s m ( R n ) , q = n p n s m p .

Together with Hölder’s inequality and Lemma 2.1, for all 0 < s < n p 1 and r = n p n s p ,

(6.2) F ( v , w ) H p s m = v m ( Δ 1 ( v w ) ) H p s m C v m ( Δ 1 ( v w ) ) L q C v L r m ( Δ 1 ( v w ) ) L n r n r C v L r m v w L r .

Now, by the Sobolev embedding theorem ( s > 0 and p ( 1 , + ) ), we have

H p s ( R n ) L r ( R n ) ,

Thus, with the estimate (6.2), we obtain

F ( v , w ) H p s m C v H p s m v w H p s .

This implies our claim.

6.2 Littlewood-Paley analysis

Let us first recall the Littlewood-Paley dyadic decomposition for a tempered distribution. Let φ 0 be a non-negative radial test function such that φ 0 ^ ( ξ ) = 1 for ξ 3 4 and such that φ 0 ^ ( ξ ) = 0 for ξ 1 . Let φ j ( x ) = 2 n j φ 0 ( 2 j x ) , i.e., φ j ^ ( x ) = φ 0 ^ ( 2 j x ) , and let us consider the partial sum operators S j associated with the φ j and defined by

S j ( f ) ( x ) = φ j f ( x ) .

Now, define ψ 0 ( x ) = φ 0 and ψ j ( x ) = φ j + 1 ( x ) φ j ( x ) and, in the same way as previously, consider the operators Δ j defined by

Δ j ( f ) ( x ) = S j + 1 ( f ) ( x ) S j ( f ) ( x ) = ψ j f ( x ) .

Thus,

(6.3) f = lim j S j ( f ) = Δ 0 + j = 0 Δ j ( f ) .

More precisely, one can prove the following result [39].

Proposition 6.1

The convergence in (6.3) occurs in H p s ( R n ) for all p in ( 1 , + ) and for all s in R . Furthermore,

f H p s Δ 0 ( f ) L p + j = 0 4 j s Δ j ( f ) 2 1 2 L p , f H p s ( R n ) .

Now, we give some classical lemmas which will be of great use in the sequel.

Lemma 6.1

[40] (Bernstein’s inequalities) Let p [ 1 , ] .

  1. If f has its spectrum in the ball B ( 0 , r ) , then there exists a constant C independent of f and r such that

    Λ s f L p C r s f L p , s > 0 .

  2. If f has its spectrum in the ring C ( 0 , A r , B r ) = { ξ : A r ξ B r } , then there exists some constants C 1 and C 2 independent of f and r such that

    C 1 r s f L p Λ s f L p C 2 r s f L p , s > 0 .

For the behavior of S j ( u ) and Δ j ( u ) in L ( R n ) -norm when u H p s ( R n ) , we give the following.

Lemma 6.2

[40] Let s n = s n p .

  1. For all s R , Δ k ( u ) L C 2 k s n u H p s ;

  2. If s < n p , then S k ( u ) L C 2 k s n u H p s .

Lemma 6.3

[32] Let { f k } k = 0 be a sequence of functions in S ( R n ) such that

supp ( f ˆ k ) B ( 0 , C 2 k ) .

Then, there exists a constant C such that

j = 0 k = 0 Δ j ( f k ) 2 1 2 L p C k = 0 f k 2 1 2 L p .

6.3 Paracomposition formula

To prove Theorem 1.1, we use the paracomposition technique (see [32,4042]), which generalizes the paraproduct technique introduced by Bony. We rewrite F ( v , w ) as the series

F ( v , w ) = F ( S 0 ( v ) , S 0 ( w ) ) + ( F ( S 1 ( v ) , S 1 ( w ) ) F ( S 0 ( v ) , S 0 ( w ) ) ) + + ( F ( S k + 1 ( v ) , S k + 1 ( w ) ) F ( S k ( v ) , S k ( w ) ) ) + = F ( S 0 ( v ) , S 0 ( w ) ) + k = 0 S k + 1 ( v ) m ( Δ ) 1 Δ k ( v w ) + k = 0 ( S k + 1 ( v ) m S k ( v ) m ) ( Δ ) 1 S k ( v w ) = F ( S 0 ( v ) , S 0 ( w ) ) + k = 0 S k + 1 ( v ) m ( Δ ) 1 Δ k ( v w ) + k = 0 Δ k ( v ) m k ( v , w ) ,

where

m k ( v , w ) = m 0 1 ( S k ( v ) + t Δ k ( v ) ) m 1 ( Δ ) 1 S k ( v w ) d t .

Thus, we have

(6.4) m k ( v , w ) L m 0 1 S k ( v ) + t Δ k ( v ) L m 1 S k + 1 ( ( Δ ) 1 S k ( v w ) ) d t C 2 k s n p ( m 1 ) v H p s m 1 2 k ( s + 1 n p ) ( Δ ) 1 ( v w ) H p s + 1 C 2 k s n p ( m 1 ) v H p s m 1 2 k ( s + 1 n p ) v w H p s .

To relocate the m k ( u ) spectrums, we introduce a second Littlewood Paley’s partition of unity

φ 1 ^ ξ A 2 k + p = 0 ψ ˆ ξ A 2 k + p = 1 ,

and m k ( v , w ) = m k , 1 ( v , w ) + p = 0 m k , p ( v , w ) , where

m k , 1 ( v , w ) = 1 φ 1 ^ ξ A 2 k m k ( v , w ) , m k , p ( v , w ) = ψ ˆ ξ A 2 k + p m k ( v , w ) .

Thus,

k = 0 Δ k ( v ) m k ( v , w ) = k = 0 Δ k ( v ) m k , 1 ( v , w ) + k = 0 p = 0 Δ k ( v ) m k , p ( v , w ) ,

and we want to prove that each of these terms belongs to H p s m ( R n ) , where s m > 0 .

We now describe the fact that the series k = 0 Δ k ( v ) m k , 1 ( v , w ) belongs to H p s m . We first give the following lemma.

Lemma 6.4

Under (H2), we have

m k , 1 ( v , w ) L C 2 k s n p ( m 1 ) v H p s m 1 2 k ( s + 1 n p ) v w H p s , k N .

Proof

Since φ 0 be a non-negative radial test function, we have

m k , 1 ( v , w ) L 1 φ 1 ^ ξ A 2 k m k ( v , w ) L 1 φ 1 ^ ξ A 2 k L 1 m k ( v , w ) L C m k ( v , w ) L C 2 k s n p ( m 1 ) v H p s m 1 2 k ( s + 1 n p ) v w H p s .

Together with (6.4) and Lemma 6.2, Lemma 6.4 is concluded.□

To prove that the series belongs to H p s m ( R n ) , it is then sufficient to show that the function

σ ( x ) = j = 0 4 j s m Δ j k = 0 Δ k ( v ) m k , 1 ( v , w ) 2 1 2 L p .

By construction, the m k , 1 ( v , w ) spectrums are in the ball B ( 0 , A 2 k ) and the Δ k ( v ) spectrums are in the rings C ( 0 , 2 1 A 2 k , 2 A 2 k ) . Taking A = 100 (for instance), then the composition m k , 1 ( v , w ) Δ k ( v ) spectrums are in some extended balls B ( 0 , A 2 k ) and so, there exists an integer N such that Δ j ( Δ k ( v ) m k , 1 ( v , w ) ) are disjointed. So,

Δ j k = 0 Δ k ( v ) m k , 1 ( v , w ) 2 = Δ j k = j N Δ k ( v ) m k , 1 ( v , w ) 2 C 4 j s m k = j N 4 k s m Δ j ( Δ k ( v ) m k , 1 ( v , w ) ) 2 ,

by Cauchy-Schwartz inequality applied to the sequences

{ 2 k s m I k j N } and { 2 k s m Δ j ( Δ k ( v ) m k , 1 ( v , w ) ) I k j N }

(note that s m > 0 is needed). Then, by definition of σ ( x ) , we obtain

σ ( x ) C j = 0 k = 0 Δ j ( 2 k s m Δ k ( v ) m k , 1 ( v , w ) ) 2 1 2 ,

and Lemma 6.3 applied to the sequence { 2 k s m I k j N } leads to

σ ( x ) L p C k = 0 4 k s m Δ k ( v ) m k , 1 ( v , w ) 2 1 2 L p .

Now, using Lemma 6.4,

Δ k ( v ) m k , 1 ( v , w ) 2 Δ k ( v ) 2 Δ k ( v ) m k , 1 ( v , w ) L 2 C 4 k s n p ( m 1 ) v H p s 2 ( m 1 ) 4 k s + 1 n p v w H p s 2 Δ k ( v ) 2 ,

and so,

σ ( x ) L p C v H p s m 1 v w H p s k = 0 4 k s Δ k ( v ) 2 1 2 L p C v H p s m v w H p s .

Thus, the series belongs to H p s m ( R n ) and its norm is bounded by C v H p s m v w H p s .

Then, we describe the fact that the series k = 0 p = 0 Δ k ( v ) m k , p ( v , w ) belongs to H p s m ( R n ) .

For the fixed p 0 , we define

p ( x ) = p = 0 Δ k ( v ) m k , p ( v , w ) .

Taking the constant A large enough, one can check that the Δ k ( v ) m k , p ( v , w ) spectrums are in some rings { ξ : C 1 2 p + k ξ C 2 2 p + k } . So, there exists an integer K (which does not depend on p ) such that those rings are K to K disjointed. So, we can use the Littlewood-Paley analysis on the K partial sums p r ( x ) defined by

p r ( x ) = k = r mod ( K ) Δ k ( v ) m k , p ( v , w ) , r { 0 , , K 1 } ,

and, by Proposition 6.1, we know that for all r { 0 , , K 1 } ,

p r H p s m C k = r mod ( K ) 4 ( k + p ) s m Δ k ( v ) m k , p ( v , w ) 2 1 2 L p .

Lemma 6.5

Under(H2), we have

m k , p ( v , w ) L C 2 k s n p m k 2 p m v H p s m 1 v w H p s .

Proof

Let us define m = N + ν , where N = [ m ] and ν [ 0 , 1 ) , and

P k t ( v ) ( x ) = S k ( v ) ( x ) + t Δ k ( v ) ( x ) .

By Lemma 6.1 applied with p = ,

(6.5) m k , p ( v , w ) L C 2 ( k + p ) m m k ( v , w ) C m ,

where C θ ( R n ) denotes the Hölder space of order θ endowed with the norm:

h C θ = h L + + D θ h L , if θ N , h C θ = h C N + sup x y < 1 D N h ( x ) D N ( y ) x y ν , if θ N .

So, by (6.5),

(6.6) m k , p ( v , w ) L C 2 ( k + p ) m m k ( v , w ) L + + D N m k ( v , w ) L + sup x y < 1 D N m k ( v , w ) ( x ) D N m k ( v , w ) ( y ) x y ν .

The bound of m k ( v , w ) L is easy to establish: we have just to argue as in the proof of Lemma 6.4 to obtain

(6.7) m k ( v , w ) L C 2 k s n p ( m 1 ) v H p s m 1 2 k ( s + 1 n p ) v w H p s .

Next, we must bound D j m k ( v , w ) L for j { 1 , , N } . Let γ be multi-index such that γ = γ 1 + + γ m with total length γ = γ 1 + + γ m = j , then,

γ m k ( v , w ) ( x ) = γ + η = γ 0 1 q = 1 γ γ 1 + + γ q = γ D q + 1 ( P k t ( x ) ) m γ 1 P k t ( x ) γ q P k t ( x ) η ( ( Δ ) 1 S k ( v w ) ) ( x ) d t .

By Lemmas 6.4 and 6.5,

(6.8) D q + 1 ( P k t ( x ) ) m L C P k t ( x ) L m q 1 C 2 k s n p ( m q 1 ) v H p s m q 1 ,

(6.9) γ i P k t ( x ) L C 2 γ i k P k t ( x ) L C 2 γ i k 2 k s n p v H p s ,

and furthermore,

(6.10) η ( ( Δ ) 1 S k ( v w ) ) L C 2 η k S k ( ( Δ ) 1 ( v w ) ) L C 2 η k 2 k ( s + 1 n p ) ( Δ ) 1 ( v w ) H p s + 1 C 2 η k 2 k ( s + 1 n p ) v w H p s .

Thus, for j { 1 , , N } ,

(6.11) D j m k ( v , w ) L C 2 j k 2 k s n p m k v H p s m 1 v w H p s .

To conclude, we must estimate

sup x y < 1 D N m k ( v , w ) ( x ) D N m k ( v , w ) ( y ) x y ν D ( x , y ) .

Let γ be a multi-index of length N . Then,

γ m k ( v , w ) ( x ) γ m k ( v , w ) ( y ) I ( x , y ) + J ( x , y ) + K ( x , y ) ,

where

(6.12) I ( x , y ) = γ + η = γ 0 1 q = 1 γ γ 1 + + γ q = γ [ D q + 1 ( P k t ( x ) ) m D q + 1 ( P k t ( y ) ) m ] i = 1 q γ i P k t ( x ) η ( ( Δ ) 1 S k ( v w ) ) ( x ) d t ,

(6.13) J ( x , y ) = γ + η = γ 0 1 q = 1 γ γ 1 + + γ q = γ D q + 1 ( P k t ( y ) ) m j = 1 q ( i > j γ i P k t ( x ) ) ( γ j P k t ( x ) γ j P k t ( y ) ) × ( i < j γ i P k t ( y ) ) η ( ( Δ ) 1 S k ( v w ) ) ( x ) d t ,

(6.14) K ( x , y ) = γ + η = γ 0 1 q = 1 γ γ 1 + + γ q = γ D q + 1 ( P k t ( y ) ) m i = 1 q γ i P k t ( y ) η ( Δ ) 1 ( S k ( v w ) ( x ) S k ( v w ) ( y ) ) d t .

By definition of C s ( R ) and Lemma 6.2, we have

(6.15) sup x y < 1 D q + 1 ( P k t ( x ) ) m D q + 1 ( P k t ( y ) ) m x y ν sup x y < 1 C ( P k t ( x ) ) m q 1 ( P k t ( y ) ) m q 1 x y ν C sup x y < 1 P k t ( x ) P k t ( y ) x y ν ( P k t ( x ) L m q 2 + P k t ( y ) L m q 2 ) C P k t C ν P k t L m q 2 C 2 k ν 2 k s n p ( m q 1 ) v H p s m q 1 ,

and

(6.16) sup x y < 1 γ j ( P k t ( v ) ( x ) P k t ( v ) ( y ) ) x y ν P k t ( v ) C γ j + ν C 2 k ( γ j + ν ) P k t ( v ) L C 2 k ( γ j + ν ) 2 k s n p v H p s .

Similarly, we have

(6.17) sup x y < 1 η ( Δ ) 1 ( S k ( v w ) ( x ) S k ( v w ) ( y ) ) x y ν C 2 k ( γ η + ν ) 2 k ( s + 1 n p ) v w H p s .

Then, by (6.8)–(6.10) and (6.12)–(6.17), we have

(6.18) D ( x , y ) = sup x y < 1 I ( x , y ) + J ( x , y ) + K ( x , y ) x y ν C 2 k m 2 k s n p m k v H p s m 1 v w H p s .

Now, by (6.6), (6.7), (6.11), and (6.18), we have

m k , p ( v , w ) L C 2 ( k + p ) m j = 0 N 2 k j 2 k s n p m k + 2 k m 2 k s n p m k v H p s m 1 v w H p s C 2 p m 2 k s n p m k j = 0 N 2 k j m k + 1 v H p s m 1 v w H p s C 2 p m 2 k s n p m k v H p s m 1 v w H p s .

This ends the proof of Lemma 6.5.□

Then, by Lemma 6.5,

p r H p s m C k = r mod ( K ) 4 ( k + p ) s m Δ k ( v ) 2 m k , p ( v , w ) L 2 1 2 L p C k = r mod ( K ) 4 k s m Δ k ( v ) 2 4 k s n p m k 1 2 L p 2 p m 2 p s m v H p s m 1 v w H p s C 2 p ( s m m ) v H p s m 1 v w H p s k = r mod ( K ) 4 k s Δ k ( v ) 2 1 2 L p C 2 p ( s m m ) v H p s m v w H p s .

Thus, for s m < m , the K series { p r } p N are uniformly convergent in H p s m ( R n ) and

r { 0 , , K 1 } p r H p s m C v H p s m v w H p s .

Thus, the series belongs to H p s m ( R n ) and its norm is bounded by C v H p s m v w H p s .

To end the proof of Theorem 1.1, we need to prove that k = 0 S k + 1 ( v ) m ( Δ ) 1 Δ k ( v w ) and F ( S 0 ( v ) , S 0 ( w ) ) belong to H p s m ( R n ) ; using the same methods as in Section 6.3, we can obtain that they are bounded by C v H p s m v w H p s .

Acknowledgements

We thank the editor and referees for their valuable comments and contributions.

  1. Funding information: The research was supported by the National Natural Science Foundation of China (No. 12171442). The research of C. Gu was partially supported by the CSC under Grant No. 202006160118.

  2. Author contributions: Caihong Gu carried out the fractional Laplacian and Yanbin Tang carried out the reaction diffusion equations. All authors carried out the proofs and conceived the study. All authors read and approved the final manuscript.

  3. Conflict of interest: The authors have no conflicts to disclose.

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

References

[1] N. Jacob, Pseudo Differential Operators and Markov Processes, 143 Vol. III: Markov Processes and Applications. Imperial College Press, London, UK, 2005. 10.1142/9781860947155Search in Google Scholar

[2] S. Chandrasekhar, Principles of Stellar Dynamics, University of Chicago Press, Chicago, Illinois, 1942. Search in Google Scholar

[3] P.-H. Chavanis, J. Sommeria, and R. Robert, Statistical mechanics of two-dimensional vortices and collisionless stellar systems, Astrophys. J. 471 (1996), 385–399. 10.1086/177977Search in Google Scholar

[4] T. Hillen and K. J. Painter, A user’s guide to PDE models for chemotaxis, J. Math. Biol. 58 (2009), 183–217. 10.1007/s00285-008-0201-3Search in Google Scholar PubMed

[5] P. Biler and G. Karch, Blowup of solutions to generalized Keller-Segel model, J. Evol. Equ. 10 (2010), 247–262. 10.1007/s00028-009-0048-0Search in Google Scholar

[6] P. Biler and G. Wu, Two dimensional chemotaxis models with fractional diffusion. Math. Meth. Appl. Sci. 32 (2009), 112–126. 10.1002/mma.1036Search in Google Scholar

[7] P. Biler, A. Boritchev, G. Karch, and P. Laurencot, Concentration phenomena in a diffusive aggregation model, J. Differ. Equ. 271 (2021), 1092–1108. 10.1016/j.jde.2020.09.035Search in Google Scholar

[8] P. Biler, A. Boritchev, G. Karch, and P. Laurencot, Sharp Sobolev estimates for concentration of solutions to an aggregation-diffusion equation, J. Dyn. Diff. Equat. 34 (2021), 3131–3141. 10.1007/s10884-021-09998-wSearch in Google Scholar

[9] P. Biler and W. A. Woyczynski, Global and exploding solutions for nonlocal quadratic evolution problems, SIAM J. Appl. Math. 59 (1998), no. 3, 845–869. 10.1137/S0036139996313447Search in Google Scholar

[10] E. F. Keller and L. A. Segel, Initiation of slide mold aggregation viewed as an instability, J. Theor. Biol. 26 (1970), 399–415. 10.1016/0022-5193(70)90092-5Search in Google Scholar PubMed

[11] E. F. Keller and L. A. Segel, Model for chemotaxis, J. Theor. Biol. 30 (1971), 225–234. 10.1016/0022-5193(71)90050-6Search in Google Scholar PubMed

[12] H. Ma and Y. Tang, Homogenization of a semilinear elliptic problem in a thin composite domain with an imperfect interface, Math. Meth. Appl. Sci. 46 (2023), no. 18, 19329–19350. 10.1002/mma.9628Search in Google Scholar

[13] Y. Zhao and Y. Tang, Critical behavior of a semilinear time fractional diffusion equation with forcing term depending on time and space, Chaos, Solitons and Fractals, 178 (2024), 114309. 10.1016/j.chaos.2023.114309Search in Google Scholar

[14] T. Ogawa and S. Shimizu, End point maximal regularity and wellposedness of the two dimensional Keller-Segel system in a critical Besov space, Math. Z. 264 (2010), 601–628. 10.1007/s00209-009-0481-3Search in Google Scholar

[15] C. Escudero, The fractional Keller-Segel model, Nonlinearity 19 (2006), 2909–2918. 10.1088/0951-7715/19/12/010Search in Google Scholar

[16] C. Gu and Y. Tang, Global solution to the Cauchy problem of fractional drift diffusion system with power-law nonlinearity, Networks Heterogeneous Media 18 (2023), 109–139. 10.3934/nhm.2023005Search in Google Scholar

[17] J. Zhao and Q. Liu, On the Cauchy problem for the fractional drift diffusion system in critical Besov spaces, Appl. Anal. 93 (2014), no. 7, 1431–1450. 10.1080/00036811.2013.833608Search in Google Scholar

[18] G. Wu and X. Zheng, On the well posedness for Keller-Segel system with fractional diffusion, Math. Meth. Appl. Sci. 34 (2011), 1739–1750. 10.1002/mma.1480Search in Google Scholar

[19] N. Bellomo, A. Bellouquid, and N. Chouhad, From a multiscale derivation of nonlinear cross diffusion models to Keller Segel models in a Navier-Stokes fluid, Math. Mod. Methods Appl. Sci. 26 (2016), no. 11, 2041–2069. 10.1142/S0218202516400078Search in Google Scholar

[20] T. Ogawa and M. Yamamoto, Asymptotic behavior of solutions to drift diffusion system with generalized dissipation. Math. Mod. Methods Appl. Sci. 19 (2009), no. 6, 939–967. 10.1142/S021820250900367XSearch in Google Scholar

[21] P. Biler, Singularities of Solutions to Chemotaxis Systems. De Gruyter Series in Mathematics and Life Sciences, vol. 6, De Gruyter, Berlin, 2020. 10.1515/9783110599534Search in Google Scholar

[22] N. Ben Abdallah, F. Mehats, and N. Vauchelet, A note on the long time behavior for the drift diffusion Poisson system, C. R. Math. Acad. Sci. Paris 339 (2004), 683–688. 10.1016/j.crma.2004.09.025Search in Google Scholar

[23] P. Biler and J. Dolbeault, Long time behavior of solutions to Nernst-Planck and Debye-Huckel drift diffusion systems. Ann. Henri Poincare. 1 (2000), 461–472. 10.1007/s000230050003Search in Google Scholar

[24] C. Gu and Y. Tang, Chaotic characterization of one dimensional stochastic fractional heat equation, Chaos, Solitons and Fractals 145 (2021), no. 5–6, 110780. 10.1016/j.chaos.2021.110780Search in Google Scholar

[25] Z. Wang, K. Zhao, P. Li, and Y. Liu, Boundedness of square functions related with fractional Schrödinger semigroups on stratified Lie groups, Commun. Anal. Mechanics 15 (2023), no. 3, 410–435. 10.3934/cam.2023020Search in Google Scholar

[26] M. K. Hamdani, L. Mbarki, and M. Allaoui, A new class of multiple nonlocal problems with two parameters and variable-order fractional p(⋅) -Laplacian, Commun. Anal. Mech. 15 (2023), no. 3, 551–574. 10.3934/cam.2023027Search in Google Scholar

[27] C. Ju, G. M. Bisci, and B. Zhang, On sequences of homoclinic solutions for fractional discrete p-Laplacian equations, Commun. Anal. Mech. 15 (2023), no. 4, 586–597. 10.3934/cam.2023029Search in Google Scholar

[28] X. Yang and Y. Tang, Decay estimates of nonlocal diffusion equations in some particle systems, J. Math. Phys. 60 (2019), no. 4, 043302. 10.1063/1.5085894Search in Google Scholar

[29] J. Chen and Y. Tang, Homogenization of nonlocal nonlinear p-Laplacian equation with variable index and periodic structure, J. Math. Phys. 64 (2023), no. 6, 061502. 10.1063/5.0091156Search in Google Scholar

[30] J. Chen and Y. Tang, Homogenization of nonlinear nonlocal diffusion equation with periodic and stationary structure, Netw. Heterogen. Media 18 (2023), no. 3, 1118–1177. 10.3934/nhm.2023049Search in Google Scholar

[31] Y. Zhao and Y. Tang, Approximation of solutions to integro-differential time fractional wave equations in Lp-space, Netw. Heterogen. Media 18 (2023), no. 3, 1024–1058. 10.3934/nhm.2023045Search in Google Scholar

[32] Y. Meyer, Remarques sur un théoréme de J. Delsarte, Annales de laInstitut Fourier 26 (1976), no. 2, 133–152. 10.5802/aif.618Search in Google Scholar

[33] F. B. Weissler, Existence and non-existence of global solutions for a semilinear heat equation, Israel J. Math. 38 (1981), 29–40. 10.1007/BF02761845Search in Google Scholar

[34] T. Kato, Strong Lp solutions of the Navier-Stokes equations in Rm with applications, Math. Z. 187 (1984), 471–480. 10.1007/BF01174182Search in Google Scholar

[35] Y. Giga and O. Sawada, On regularizing-decay rate estimates for solutions to the Navier-Stokes initial value problem, Anal. Appl. 1 (2003), no. 2, 549–562. Search in Google Scholar

[36] P. G. Lemarie-Rieusset, Recent Developments in the Navier-Stokes Problem, Chapman and Hall/CRC Research Notes in Mathematics Series, CRC Press, UK, 2002. 10.1201/9780367801656Search in Google Scholar

[37] W. P. Ziemer, Weakly Differentiable Functions, Graduate Texts in Mathematics, vol. 120, Springer Verlag, 1989. 10.1007/978-1-4612-1015-3Search in Google Scholar

[38] G. Wu and J. Yuan, Well-posedness of the Cauchy problem for the fractional power dissipative equation in critical Besov spaces, J. Math. Anal. Appl. 340 (2008), 1326–1335. 10.1016/j.jmaa.2007.09.060Search in Google Scholar

[39] H. Triebel, Theory of Function Spaces, Monographs in Mathematics, vol. 78, Birkhauser Verlag, Basel, 1983. 10.1007/978-3-0346-0416-1Search in Google Scholar

[40] S. Alinhac and P. Gérard, Operateurs pseudo-differential et théoréme de Nash-Moser, Savoirs Actuels. InterEditions, Paris, 1991. 10.1051/978-2-7598-0282-1Search in Google Scholar

[41] M. E. Taylor, Pseudo differential operators and nonlinear PDE, Birkhauser Verlag, Basel, 1991. 10.1007/978-1-4612-0431-2Search in Google Scholar

[42] T. Colin, On the Cauchy problem for dispersive equations with nonlinear terms involving high derivatives and with arbitrarily large initial data. Nonlinear Anal. 22 (1994), 835–845. 10.1016/0362-546X(94)90052-3Search in Google Scholar

Received: 2023-04-18
Revised: 2023-11-02
Accepted: 2024-06-03
Published Online: 2024-08-05

© 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-2024-0023/html
Scroll to top button