Loading [MathJax]/jax/output/SVG/jax.js
Research article Special Issues

Normalized solutions for the mixed dispersion nonlinear Schrödinger equations with four types of potentials and mass subcritical growth

  • Received: 12 December 2022 Revised: 15 March 2023 Accepted: 27 March 2023 Published: 28 April 2023
  • This paper is devoted to considering the attainability of minimizers of the L2-constraint variational problem

    mγ,a=inf{Jγ(u):uH2(RN),RN|u|2dx=a2},

    where

    Jγ(u)=γ2RN|Δu|2dx+12RN|u|2dx+12RNV(x)|u|2dx12σ+2RN|u|2σ+2dx,

    γ>0, a>0, σ(0,2N) with N2. Moreover, the function V:RN[0,+) is continuous and bounded. By using the variational methods, we can prove that, when V satisfies four different assumptions, mγ,a are all achieved.

    Citation: Cheng Ma. Normalized solutions for the mixed dispersion nonlinear Schrödinger equations with four types of potentials and mass subcritical growth[J]. Electronic Research Archive, 2023, 31(7): 3759-3775. doi: 10.3934/era.2023191

    Related Papers:

    [1] Pierre Gabriel, Hugo Martin . Steady distribution of the incremental model for bacteria proliferation. Networks and Heterogeneous Media, 2019, 14(1): 149-171. doi: 10.3934/nhm.2019008
    [2] Fadia Bekkal-Brikci, Giovanna Chiorino, Khalid Boushaba . G1/S transition and cell population dynamics. Networks and Heterogeneous Media, 2009, 4(1): 67-90. doi: 10.3934/nhm.2009.4.67
    [3] Pierre Degond, Sophie Hecht, Nicolas Vauchelet . Incompressible limit of a continuum model of tissue growth for two cell populations. Networks and Heterogeneous Media, 2020, 15(1): 57-85. doi: 10.3934/nhm.2020003
    [4] Paulo Amorim, Alessandro Margheri, Carlota Rebelo . Modeling disease awareness and variable susceptibility with a structured epidemic model. Networks and Heterogeneous Media, 2024, 19(1): 262-290. doi: 10.3934/nhm.2024012
    [5] Piotr Gwiazda, Karolina Kropielnicka, Anna Marciniak-Czochra . The Escalator Boxcar Train method for a system of age-structured equations. Networks and Heterogeneous Media, 2016, 11(1): 123-143. doi: 10.3934/nhm.2016.11.123
    [6] Martino Bardi . Explicit solutions of some linear-quadratic mean field games. Networks and Heterogeneous Media, 2012, 7(2): 243-261. doi: 10.3934/nhm.2012.7.243
    [7] Peter V. Gordon, Cyrill B. Muratov . Self-similarity and long-time behavior of solutions of the diffusion equation with nonlinear absorption and a boundary source. Networks and Heterogeneous Media, 2012, 7(4): 767-780. doi: 10.3934/nhm.2012.7.767
    [8] Yacine Chitour, Guilherme Mazanti, Mario Sigalotti . Stability of non-autonomous difference equations with applications to transport and wave propagation on networks. Networks and Heterogeneous Media, 2016, 11(4): 563-601. doi: 10.3934/nhm.2016010
    [9] Dieter Armbruster, Christian Ringhofer, Andrea Thatcher . A kinetic model for an agent based market simulation. Networks and Heterogeneous Media, 2015, 10(3): 527-542. doi: 10.3934/nhm.2015.10.527
    [10] Andrea Tosin . Multiphase modeling and qualitative analysis of the growth of tumor cords. Networks and Heterogeneous Media, 2008, 3(1): 43-83. doi: 10.3934/nhm.2008.3.43
  • This paper is devoted to considering the attainability of minimizers of the L2-constraint variational problem

    mγ,a=inf{Jγ(u):uH2(RN),RN|u|2dx=a2},

    where

    Jγ(u)=γ2RN|Δu|2dx+12RN|u|2dx+12RNV(x)|u|2dx12σ+2RN|u|2σ+2dx,

    γ>0, a>0, σ(0,2N) with N2. Moreover, the function V:RN[0,+) is continuous and bounded. By using the variational methods, we can prove that, when V satisfies four different assumptions, mγ,a are all achieved.



    In structured population dynamics, finding the structuring variable(s) which best describes a phenomenon is a crucial question. For a population of proliferating cells or bacteria the variables usually considered are age, size (see [27,12,18]) or a combination of both (see [1,24,10] for modeling and [26,12,10,6] for mathematical analysis). Recent experimental work highlighted the limits of these models to describe bacteria, and a new variable to trigger division emerged: the size-increment, namely the size gained since the birth of the cell (see [22] and references therein for a review of the genesis of the related model). This so called 'adder principle ensures homeostasis with no feedback from the bacteria and explains many experimental data. In this model, bacteria are described by two parameters: their size-increment and their size, respectively denoted by a and x in the following (the choice of letter a is reminiscent from the age variable, since as for the age, the size increment is reset to zero after division). This choice of variables is motivated by the main assumption of the model, which is that the control of the cellular reproduction is provided by the division rate B which is supposed to depend only on a, and the growth rate g which is assumed to depend only on x. With the variables we introduced, the model formulated in [25] reads

    {tn(t,a,x)+a(g(x)n(t,a,x))+x(g(x)n(t,a,x))+B(a)g(x)n(t,a,x)=0,t0, x>a>0,g(x)n(t,0,x)=4g(2x)0B(a)n(t,a,2x)da,t0, x>0.

    The function n(t,a,x) represents the number of cells at time t of size x that have grown of an increment a since their birth. The boundary term denotes an equal mitosis, meaning that after division, a mother cell gives birth to two daughters of equal size. However, if this special case of equal mitosis is appropriate to describe the division of some bacterium (e.g. E. Coli), it is inadequate for asymetric division (like yeast for instance) or for a fragmentation involving more than two daughters (as in the original model formulated for plant growth in [10]). In the current paper, we propose to consider more general division kernels. We assume that when a cell of size x divides, it gives birth to a daughter of size zx with a certain probability which depends on z(0,1) but is independent of x. Such fragmentation process is usually called self-similar. More precisely the number of daughters with a size between zx and (z+dz)x is given by μ([z,z+dz]), where μ is a positive measure on [0,1]. The model we consider is then formulated as

    tn(t,a,x)+a(g(x)n(t,a,x))+x(g(x)n(t,a,x))+B(a)g(x)n(t,a,x)=0,t0, x>a>0, (1a)
    g(x)n(t,0,x)=10g(xz)0B(a)n(t,a,xz)da dμ(z)z,t0, x>0. (1b)

    It appears that this model is a particular case of the one proposed in the pioneer work [10] for plants growing in a single dimension, mixing age and size control. Indeed, in this paper the authors noticed that in the case of a deterministic and positive growth rate, a size/age model is equivalent to a size/birth-size through the relation a=xs, where s denotes the birth-size (see Figure 1). They preferred working with the size/birth-size description since in this framework the transport term acts only in the x direction. In the case when g is independent of x and B is bounded from above and below by positive constants, it is proved in [26] for μ a uniform measure on [0,1], and in [12,Chapter V] for the equal mitosis, that the solutions to the system (1) converge to a stable distribution as time goes to infinity. In the present paper we propose to study the model (1) in the case of a linear growth rate (see [1] for a discussion on this hypothesis). More precisely we are interested in populations which evolve with a stable size and size-increment distribution, i.e. solutions of the form n(t,a,x)=h(t)N(a,x). The existence of such separable solutions when g is linear was already the topic of [10], but their proof required the equation to be set on a bounded domain and they had to impose a priori the existence of a maximal size for the population. In our case no maximal size is prescribed and it brings additional difficulties due to a lack of compactness. To address this problem, we will make the following assumptions.

    Figure 1. 

    schematic representation of the variables on an E. coli bacterium

    .

    First, we want the sum of the daughters sizes to be equal to the size of the mother. This rule, called mass conservation, prescribes

    10zdμ(z)=1. (2)

    We also assume that the division does not produce any arbitrarily small daughter by imposing that the support of μ is a compact subset of (0,1), which ensures that

    θ:=infsuppμ>0andη(θ,1), suppμ[θ,η]. (3)

    In particular, these assumptions imply that the mean number of daughters μ([0,1]) is finite. The division rate B is assumed to be a nonnegative and locally integrable function on R+ such that

    b0,suppB=[b,), (4)

    see [8] for instance. It will be useful in our study to define the associated survivor function Ψ by

    Ψ(a)=ea0B(z)dz.

    For a given increment a, Ψ(a) represents the probability that a cell did not divide before having grown of at least a since its birth. We assume that the function B is chosen in such a way that Ψ tends to zero at infinity, meaning that all the cells divide at some time. More precisely we make the following quantitative assumption

    k0>0,Ψ(a)=+O(ak0). (5)

    This assumption on the decay at infinity of the survivor function enables a wide variety of division rates. For instance, it is satisfied if there exists A>0 such that

    aA,B(a)k0a.

    The function B being locally integrable, the function Ψ belongs to W1,1loc(R+) and (5) ensures that its derivative belongs to L1(R+). We can introduce the useful function Φ defined by

    Φ=BΨ=Ψ (6)

    which is the probability distribution that a cell divides at increment a. Recall that, as in [10], we consider the special case of a linear growth rate, namely g(x)=x. In this case, multiplying by the size x and integrating, we obtain ddtxn(t,a,x)dadx=xn(t,a,x)dadx, and so

    xn(t,a,x)dadx=etxn0(a,x)dadx. (7)

    This implies that if we look for a solution with separated variables n(t,a,x)=h(t)N(a,x), necessarily h(t)=h(0)et. In other words, the Malthus parameter of the population is 1. This motivates the Perron problem which consists in finding N=N(a,x), which is solution to the system

    a(xN(a,x))+x(xN(a,x))+(1+xB(a))N(a,x)=0,x>a>0, (8a)
    N(0,x)=100B(a)N(a,xz)da dμ(z)z2,x>0, (8b)
    N(a,x)0,xa0, (8c)
    0x0N(a,x)dadx=1. (8d)

    It is convenient to define the set Ω0:={(a,x)R2, 0ax}, and we are now ready to state the main result of the paper.

    Theorem 1.1. Let μ be a positive measure on [0,1] satisfying (2) and (3), and B be a nonnegative and locally integrable function on R+ satisfying (4) such that the associated survivor function Ψ satisfies (5). Then, there exists a unique solution NL1(X,(1+(xa)2)dadx) to the eigenproblem (8). This solution is expressed as

    N:(a,x)XΨ(a)x2f(xa) (9)

    where f is a nonnegative function which satisfies

    fL1(R+,xldx)

    for all l<k0, k0 being the positive number given in hypothesis (5), and

    suppf=[bθ,)

    with bθ=θ1θb, where θ and b are defined in (3) and (4) respectively.

    The fast decay of the function f near zero is a consequence of the form of the support of the fragmentation kernel μ. Furthermore, this decay is consistent with the decay near zero of the eigenvector for the size equation (see [8]). Remark that for any nonnegative and appropriately normalized function fL1(R+), the expression given in (9) satisfies (8a), (8c), and (8d). The proof of Theorem 1.1 consists in finding the appropriate function f such that (8b) is also satisfied. This function is obtained as the fixed point of a conservative operator, and this allows us to compute it numerically by using the power iteration (see [20]). We obtain the function on the left on Figure 2. On the right is the related density N(a,x).

    Figure 2. 

    Left: simulation of the function f by the power method with B(a)=21+a1{1a} and μ(z)=2δ12(z). Right: level set of the density N(a,x) obtained from this function f. Straight line: the set {x=a+1}.

    .

    Notice also that for the function N given by (9), the function sN(a+s,x+s) is continuous for any ax. It corresponds to the trajectories along the characteristics.

    The article is organised as follows. In Section 2 we reduce the Perron eigenvalue problem with two variables to a fixed point problem for an integral operator in dimension one. Section 3 is dedicated to proving the existence and uniqueness of the fixed point by using functional analysis and Laplace transform methods. In Section 4 we go through the usefulness of knowing N to develop entropy methods. Finally in Section 5 we discuss some interesting perspectives.

    Our study consists in constructing a solution to the eigenproblem (8) from the solution of a fixed point problem. First, we notice that the size x of a cell and its size increment a grow at the same speed g(x), so the quantity xa remains constant: it corresponds to the birth-size of the cell, denoted by s. To simplify the equation and obtain horizontal straight lines as characteristics (see Figure 3), we give a description of the population with size increment a and birth-size s, namely we set

    M(a,s):=N(a,a+s). (10)
    Figure 3. 

    Domain of the model, with respect to the choice of variables to describe the bacterium. Grey: domain where the bacteria densities may be positive. Arrows: transport. Left: size increment/size. Right: size increment/birth size. Dashed: location of cells of size x1

    .

    Thanks to this relation, it is equivalent to prove the existence of an eigenvector for the increment-size system or for the increment/birth-size system. To determine the equation verified by M, we compute the partial derivatives of xN(a,x)=(a+s)M(a,s), which leads to the equation

    a((a+s)M(a,s))+(1+(a+s)B(a))M(a,s)=0.

    Writing the non-local boundary condition (8b) with the new variables takes less calculation and more interpretation. In (8b) the number of cells born at size s resulted of the division of cells at size sz. Then the equivalent of (8b) in the new variables with a linear growth rate is given by

    M(0,s)=ηθsz0B(a)M(a,sza)da dμ(z)z2

    since there is no mass for asz. With the relation (10), it is equivalent to solve (8) and to solve

    a((a+s)M(a,s))+(1+(a+s)B(a))M(a,s)=0,a,s>0, (11a)
    M(0,s)=ηθsz0B(a)M(a,sza)da dμ(z)z2,s>0, (11b)
    M(a,s)0,a,s0, (11c)
    R2+M(a,s)dads=1. (11d)

    Considering the variable s as a parameter in (11a), we see this equation as an ODE in the variable a. A formal solution is given by

    M(a,s)=Ψ(a)(a+s)2s2M(0,s).

    Having this expression in mind, we note that for any nonnegative function fL1(R+,ds), the function Mf defined on R2+ by

    Mf:(a,s)Ψ(a)(a+s)2f(s)

    is a solution of (11a) and satisfies (11c). Then it remains to choose the appropriate function f and normalize the related function Mf to solve the whole system (11). It turns out that this appropriate function f is a fixed point of the operator T:L1(R+)L1(R+) defined by

    Tf(s)=ηθsz0Φ(sza)f(a)dadμ(z), (12)

    where Φ=BΨ, as stated in the following lemma.

    Lemma 2.1. The function Mf satisfies (11b) if and only if f is a fixed point of the operator T.

    Proof.

    Mf satisfies (11b)f(s)s2=ηθsz0B(a)Ψ(a)(sz)2f(sza)da dμ(z)z2f(s)=ηθsz0Φ(a)f(sza)da dμ(z)f(s)=ηθsz0Φ(sza)f(a)da dμ(z)f(s)=Tf(s)

    The operator T can be seen as some kind of transition operator: it links the laws of birth size of two successive generations. If f is the law of the parents, then Tf is the law of the birth size of the newborn cells. Indeed, Equation (12) can be understood in words as 'the number of cells born at size s come from the ones that were born at size a[bθ,sz] and elongated of sza, for all z[θ,η] and all a, before dividing into new cells'. See [7] for a probabilistic viewpoint on the conservative size equation. It is easy to check that T is a continuous linear operator on L1(R+) and that TL(L1(R+))ΦL1(R+)=1 using (2) and (6). The following lemma provides a slightly stronger result.

    Lemma 2.2. For all l0, the operator T maps continuously L1(R+,slds) into itself. Additionally, if (5) holds true, then T maps continuously L1(R+,(sk+sl)ds) into itself for any l0 and k[0,k0).

    Proof. We start with L1(R+,slds) where l0. For fL1(R+,slds) and β>α>0 one has

    βα|Tf(s)|sldsηθβαslsz0Φ(sza)|f(a)|dadsdμ(z)ηθαz0|f(a)|βαΦ(sza)sldsdadμ(z)+ηθβzαz|f(a)|βzaΦ(sza)sldsdadμ(z)ηθαz0|f(a)|βzaαzaΦ(σ)(a+σ)lzl+1dσdadμ(z)+ηθβzαz|f(a)|βza0Φ(σ)(a+σ)lzl+1dσdadμ(z)ηθzl+1αz0|f(a)|aldadμ(z)+ηθzl+1αz|f(a)|aldadμ(z)θlfL1(R+,slds),

    which gives the conclusion by passing to the limits α0 and β+.

    For the second part we begin with the proof that under condition (5), for any k[0,k0) one has

    0Φ(a)akda<.

    First, recall that 0Φ(a)da=1 and Φ=Ψ. Integrating by parts for β1, one has

    β0Φ(a)akda10Φ(a)da+β1Φ(a)akda1+kβ1Ψ(a)ak1da

    and the last integral converges when β+ under Assumption (5) because k<k0. Now let l0 and k[0,k0), and let fL1(R+,(sk+sl)ds). Due to the first part of the proof, we only have to estimate β0|Tf(s)|skds for β>0. Since the function x(1+x)k1+xk is uniformly bounded on R+, there exists of a constant C>0 such that (a+σ)kC(ak+σk) for all a,σ0, and it allows us to write for any β>0

    β0|Tf(s)|skdsηθβ0sksz0Φ(sza)|f(a)|dadsdμ(z)=ηθβz0|f(a)|βza0Φ(σ)(a+σ)kzk+1dσdadμ(z)Cηθzk+1βz0|f(a)|akβza0Φ(σ)dσdadμ(z)+Cηθzk+1βz0|f(a)|βza0Φ(σ)σkdσdadμ(z)Cηk(fL1(R+,skds)+ΦL1(R+,skds)fL1(R+)).

    In this section we prove the existence of a unique nonnegative and normalized fixed point of the operator T.

    Let us first recall some definitions from the Banach lattices theory (for more details, see [9,23]). Let Ω be a subset of R+ and ν be a positive measure on Ω. The space L1(Ω,ν) is an ordered set with the partial order defined by

    f0 if and only if f(s)0 ν-a.e. on Ω.

    Furthermore, endowed with its standard norm, the space L1(Ω,ν) is a Banach lattice, i.e. a real Banach space endowed with an ordering compatible with the vector structure such that, if f,gL1(Ω,ν) and |f||g|, then fL1(Ω,ν)gL1(Ω,ν). A vector subspace IL1(Ω,ν) is called an ideal if fI,gL1(Ω,ν) and |g||f| implies gI. For a given operator A defined on L1(Ω,ν), a closed ideal I is A-invariant if A(I)I, and A is irreducible if the only A-invariant ideals are {0} and L1(Ω,ν). To each closed ideal I in the Banach lattice L1(Ω,ν) corresponds a subset ωΩ such that I={fL1(Ω,ν),suppfω}. We also define the positive cone L1+(Ω,ν):={fL1(Ω,ν) | f0 ν-a.e. on Ω}. An operator A:L1(Ω,ν)L1(Ω,ν) is said to be positive if A(L1+(Ω,ν))L1+(Ω,ν).

    To prove the existence of an eigenvector associated to the eigenvalue 1, we will use the following theorem, easily deduced from Krein-Rutman's theorem (see [9] for instance) and De Pagter's [5].

    Theorem 3.1. Let A:L1(Ω,ν)L1(Ω,ν) be a non-zero positive compact irreducible operator. Then its spectral radius ρ(A) is a nonnegative eigenvalue associated to a nonzero eigenvector belonging to the positive cone L1+(Ω,ν).

    Due to a lack of compactness of the operator T, which is due to the lack of compactness of R+, we truncate the operator T into a family of operators (TΣ)Σ. Let bθ=θ1θb and for Σ>bθ define the operator TΣ on L1((bθ,Σ)) by

    TΣf(s)=ηθmin(sz,Σ)bθΦ(sza)f(a)dadμ(z) (13)
    ={ηθszbθΦ(sza)f(a)dadμ(z),bθs<θΣ,sΣθΣbθΦ(sza)f(a)dadμ(z)+ηsΣszbθΦ(sza)f(a)dadμ(z),θΣsηΣ,ηθΣbθΦ(sza)f(a)dadμ(z),ηΣ<sΣ.

    Defining the lower bound of the domain as bθ will ensure the irreducibility of TΣ. We will apply Theorem 3.1 to the operator TΣ for Σ large enough to prove the existence of a pair (ρΣ,fΣ) such that TΣfΣ=ρΣfΣ. Then, we will prove that there exists a unique f in a suitable space such that ρΣ1 and fΣf as Σ, with f satisfying Tf=f.

    The following lemma ensures that the truncated operator TΣ is well defined.

    Lemma 3.2. If fL1(R+) and suppf[bθ,Σ], then

    (Tf)|[bθ,Σ]=TΣ(f|[bθ,Σ]).

    Lemma 3.2 is a straightforward consequence of the definition of operator TΣ by (13). From Lemmas 2.2 and 3.2, we deduce that TΣ has the same stability mapping properties as T. To prove the compactness of the operator TΣ for a fixed Σ and later on that the family (TΣ)Σ is also compact, we use a particular case of a corollary of the Riesz-Fréchet-Kolmogorov theorem. First, we define two properties for a bounded subset F of L1(Ω,ν) with Ω an open subset of R+, and ν a positive measure, the first on translations and the second on the absence of mass on the boundary of the domain

    {ϵ>0, ω⊂⊂Ω, δ(0,dist(ω,cΩ)) such that τhffL1(ω,ν)<ϵ, h(δ,δ), fF (14)
    {ϵ>0, ω⊂⊂Ω, such that fL1(Ωω,ν)<ϵ, fF (15)

    where cΩ is understood as the complement of this set in R+.

    Theorem 3.3 (from [3], corollary 4.27). If F is a bounded set of L1(Ω,ν) such that (14) and (15) hold true, then F is relatively compact in L1(Ω,ν).

    Using Theorem 3.1, we prove the existence of an eigenpair (ρΣ,fΣ) for the operator TΣ.

    Proposition 1. Let l be a nonpositive number. Under the hypotheses (2), (3) and (4), there exists a unique normalized eigenvector fΣL1+((bθ,Σ),slds) of the operator TΣ in L1((bθ,Σ),slds) associated to the spectral radius ρΣ for every Σ>max(11θb,1).

    Applying Theorem 3.3, to Ω=(bθ,Σ) and the family

    F={TΣf,fL1((bθ,Σ),slds),fL1((bθ,Σ),slds)1},

    which is bounded in L1((bθ,Σ),slds), as already shown in the proof of Lemma 2.2, we prove the following Lemma.

    Lemma 3.4. Let l be a nonpositive number. Under the hypotheses of Proposition 1, for all

    Σ>max(b1θ,1),

    the set F is relatively compact.

    Proof of Lemma 3.4. The set F is bounded due to the continuity of T proven in Lemma 2.2. First, we show that (14) is satisfied. Any compact set in (bθ,Σ) is included in a segment [α,β]. Without loss of generality, we take bθ<α<θΣ, ηΣ<β<Σ. It is sufficient to treat the case h positive, so let 0h<min(θΣα,Σβ,Σ(ηθ)). Since TΣf is piecewise defined, we have to separate the integral on [α,β] into several parts, depending on the interval s and s+h belong to, and we obtain

    βα|TΣf(s+h)TΣf(s)|sldsθΣhα|TΣf(s+h)TΣf(s)|slds=:(A)+θΣθΣh|TΣf(s+h)TΣf(s)|slds=:(B)+ηΣhθΣ|TΣf(s+h)TΣf(s)|slds=:(C)+ηΣηΣh|TΣf(s+h)TΣf(s)|slds=:(D)+βηΣ|TΣf(s+h)TΣf(s)|slds=:(E).

    since for (A), (C) and (E), TΣf and τhTΣf have the same expression, the same kind of calculations apply, so we only treat (C), which has the most complicated expression.

    (C)=ηΣhθΣ|TΣf(s+h)TΣf(s)|sldsηΣhθΣ|s+hΣθΣbθΦ(s+hza)f(a)dadμ(z)sΣθΣbθΦ(sza)f(a)dadμ(z)|slds+ηΣhθΣ|ηs+hΣs+hzbθΦ(s+hza)f(a)dadμ(z)ηsΣszbθΦ(sza)f(a)dadμ(z)|sldsηΣhθΣslsΣθΣbθ|Φ(s+hza)Φ(sza)||f(a)|dadμ(z)ds=:(C1)+ηΣhθΣsls+hΣsΣΣbθΦ(s+hza)|f(a)|dadμ(z)ds=:(C2)
    +ηΣhθΣslηs+hΣszbθ|Φ(s+hza)Φ(sza)||f(a)|dadμ(z)ds=:(C3)+ηΣhθΣslηs+hΣs+hzszΦ(s+hza)|f(a)|dadμ(z)ds=:(C4)+ηΣhθΣsls+hΣsΣszbθΦ(s+hza)|f(a)|dadμ(z)ds=:(C5)

    The integrals (C1) and (C3) are dealt with in the same way, and we have the following estimate

    (C1)=ηΣhθΣslsΣθΣbθ|Φ(s+hza)Φ(sza)||f(a)|dadμ(z)ds=ηhΣθΣbθ|f(a)|ηΣhzΣ|Φ(s+hza)Φ(sza)|sldsdadμ(z)=ηhΣθzl+1Σbθ|f(a)|ηΣhzaΣa|τhzΦ(σ)Φ(σ)|(a+σ)ldσdadμ(z)=ηhΣθzl+1Σbθ|f(a)|alηΣhzaΣa|τhzΦ(σ)Φ(σ)|dσdadμ(z)θlsupε[θ,η]τhεΦΦL1(R+).

    These integrals are as small as needed when h is small enough, due to the continuity of the translation in L1(R+). For (C2) one has

    (C2)=ηΣhθΣsls+hΣsΣΣbθΦ(s+hza)|f(a)|dadμ(z)ds=θ+hΣθΣbθ|f(a)|zΣθΣΦ(s+hza)sldsdadμ(z)+ηhΣθ+hΣΣbθ|f(a)|zΣzΣhΦ(s+hza)sldsdadμ(z)+ηηhΣΣbθ|f(a)|ηΣhzΣhΦ(s+hza)sldsdadμ(z)ηθΣbθ|f(a)|zΣzΣhΦ(s+hza)sldsdadμ(z)=ηθzl+1Σbθ|f(a)|ΣaΣhzaΦ(σ+hz)(a+σ)ldσdadμ(z)ηθzl+1Σbθ|f(a)|alΣaΣhzaΦ(σ+hz)dσdadμ(z)θlsup|I|=hθIΦ(a)da

    which is small when h is small since Φ is a probability density. To deal with (C4), we use Fubini's theorem and some changes of variables to obtain

    (C4)=ηΣhθΣslηs+hΣs+hzszΦ(s+hza)|f(a)|dadμ(z)ds=ηθ+hΣzΣhθΣsls+hzszΦ(s+hza)|f(a)|dadsdμ(z)=ηθ+hΣzΣhθΣsl0hzΦ(hz+a)|f(sza)|dadsdμ(z)=ηθ+hΣ0hzΦ(hz+a)zΣhθΣ|f(sza)|sldsdadμ(z)=ηθ+hΣzl+10hzΦ(hz+a)ΣhzaθΣza|f(σ)|(σ+a)ldσdadμ(z)ηθ+hΣzl+10hzΦ(hz+a)dadμ(z)θl(1Ψ(hθ)),

    and the continuity of Ψ at 0 provides the wanted property. Finally, noticing that (C5)(C2) because the integrand are nonnegative, we obtain the desired control on the integral (C). Now for the integral (B), which is dealt with as would be (D), we write

    (B)=θΣθΣh|s+hΣθΣbθΦ(s+hza)f(a)dadμ(z)+ηs+hΣs+hzbθΦ(s+hza)f(a)dadμ(z)ηθszbθΦ(sza)f(a)dadμ(z)|sldsηθΣbθ|f(a)|θΣθΣh[Φ(s+hza)+Φ(sza)]sldsdadμ(z)ηθzl+1Σbθ|f(a)|alθΣzaθΣhza[Φ(σ+hz)+Φ(σ)]dσdadμ(z)2θlsup|I|=hθIΦ(a)da

    and again the last term vanishes as h vanishes. We now show that there is no mass accumulation at the boundary of the domain (bθ,Σ), i.e. that (15) holds true. For Σ>11θb, we haves bθ<θΣ and we can choose α<θΣ, so that for all s(bθ,α), sΣ<θ. With the expression of TΣf(s), we have

    αbθ|TΣf(s)|slds
    αbθslηθszbθΦ(sza)|f(a)|dadμ(z)dsηθbθzbθ|f(a)|αbθΦ(sza)sldsdadμ(z)+ηθαzbθz|f(a)|αzaΦ(sza)sldsdadμ(z)ηθzl+1bθzbθ(Ψ(bθza)Ψ(αza))|f(a)|aldazdμ(z)+ηθαzbθz(1Ψ(αza))|f(a)|aldazdμ(z)θl(1Ψ(αbθθ)), (16)

    since for bθ<sbθz we have bθzab and so Ψ(bθza)=1. Taking α as closed to bθ as needed, we obtain the first estimate of (15).

    As done before, we choose a β to obtain a simpler expression of TΣ, namely β>ηΣ. Then, one has

    Σβ|TΣf(s)|sldsηθΣbθ|f(a)|ΣβΦ(sza)sldsdadμ(z)ηθzl+1Σbθ|f(a)|alΣzaβzaΦ(σ)dσdadμ(z)θlsup|I|=ΣβθIΦ(a)da, (17)

    which is small when Σβ is small.

    We have checked the assumptions of Theorem 3.3 for the family F, so it is relatively compact.

    To prove the irreducibility of the operator TΣ, it is useful to notice that TΣ can be expressed differently after switching the two integrals. One has

    TΣf(s)={sηbθf(a)ηθΦ(sza)dμ(z)da+sθsηf(a)saθΦ(sza)dμ(z)da,bθs<θΣ,sηbθf(a)ηθΦ(sza)dμ(z)da+Σsηf(a)saθΦ(sza)dμ(z)da,θΣsηΣ,Σbθf(a)ηθΦ(sza)dμ(z)da,ηΣ<sΣ.

    Lemma 3.5. Let l be a nonpositive number. Under the hypotheses of Proposition 1, for all Σ>11θb, the operator TΣ:L1((bθ,Σ),slds)L1((bθ,Σ),slds) is irreducible.

    Proof. Let J{0} be a TΣ-invariant ideal in L1((bθ,Σ),slds). There exists a subset ω(bθ,Σ) such that J={fL1((bθ,Σ),slds) | suppfω}. Let fω:=sl1ω and s0:=infsuppfωbθ. Since J{0} (so s0<Σ) and θ<η, one can find ζ and ξ both positive such that

    TΣfω(s)s0+ζs0θ+ξθΦ(sza)dμ(z)fω(a)da.

    For ss0, the functions zsza and asza are continuous decreasing functions. So, if s is such that sθs0>b, then one can choose ζ and ξ such that for all (s,a)[s0,s0+ζ]×[θ,θ+ξ], szasuppΦ. Additionally, for each ζ>0, the integral s0+ζs0fω(a)da is positive. We deduce that [θ(b+s0),Σ]suppTΣfω[s0,Σ], so s0θ(b+s0), which is equivalent to s0bθ. Finally bθ=s0, so J=[bθ,Σ] and TΣ is irreducible.

    Proof of Proposition 1. Lemma 3.4 shows that the set F is relatively compact in L1((bθ,Σ),slds), which is exactly saying that TΣ is a compact operator of the Lebesgue space L1((bθ,Σ),slds). With Lemma 3.5 in addition, we can apply Theorem 3.1 to the operator TΣ for Σ>11θb to obtain the existence of a nonnegative function fΣL1((bθ,Σ),slds) which is an eigenvector of TΣ associated to the eigenvalue ρΣ. Since this function is defined on a compact subset of R+, it also belongs to L1((bθ,Σ),(sk+sl)ds) for k<k0.

    We now want to show that up to a subsequence, (fΣ)Σ converges to a fixed point of T. To that end, in the rest of the article we extend the functions defined on (bθ,Σ) to R+ by 0 out of (bθ,Σ). Then we obtain the following proposition

    Proposition 2. Under hypotheses (2)- (5) there exists a nonnegative and normalized fixed point

    fL1(R+,(sk+sl)ds)

    for all l0 and k<k0, of the operator T. Additionally, f is unique in L1(R+) and its support is [bθ,).

    First, we will show that the sequence (ρΣ)Σ converges to 1 as Σ.

    Lemma 3.6. If (ρΣ,fΣ) is an eigenpair of the operator TΣ, then the following inequality holds true

    1Ψ((1η1)Σ)ρΣ1Ψ(Σθbθ). (18)

    Proof. Integrating the equality ρΣfΣ=TΣfΣ over (bθ,Σ), one has

    ρΣΣbθfΣ(s)ds=θΣbθηθszbθΦ(sza)f(a)dadμ(z)ds=:(A)+ηΣθΣsΣθΣbθΦ(sza)f(a)dadμ(z)ds=:(B)+ηΣθΣηsΣszbθΦ(sza)f(a)dadμ(z)ds=:(C)+ΣηΣηθΣbθΦ(sza)f(a)dadμ(z)ds=:(D)
    (A)=ηθbθzbθfΣ(a)θΣbθzΦ(sza)dsdadμ(z)+ηθθΣzbθzfΣ(a)θΣzaΦ(sza)dsdadμ(z)=ηθzbθzbθfΣ(a)[Ψ(bθza)Ψ(θΣza)]dadμ(z)+ηθzθΣzbθzfΣ(a)[1Ψ(θΣza)]dadμ(z)(B)=ηθfΣ(a)ΣbθηΣzΣΦ(sza)dsdadμ(z)=ηθzΣbθfΣ(a)[Ψ(Σa)Ψ(ηΣza)]dadμ(z)
    (C)=ηθθΣzbθfΣ(a)zΣθΣΦ(sza)dsdadμ(z)+ηθΣθΣzfΣ(a)zΣzaΦ(sza)dsdadμ(z)=ηθzθΣzbθfΣ(a)[Ψ(θΣza)Ψ(Σa)]dadμ(z)+ηθzΣθΣzfΣ(a)[1Ψ(Σa)]dadμ(z)(D)=ηθΣbθfΣ(a)ΣηΣΦ(sza)dsdadμ(z)=ηθzΣbθfΣ(a)[Ψ(ηΣza)Ψ(Σza)]dadμ(z)

    Then notice that for a(bθ,Σ) and z(θ,η) one has

    bθzabθzbθ=bθ(1z1)bθ(1θ1)=b,

    so as in the computations leading to (16), Ψ(bθza)=1. Combining these different expressions, we deduce

    ρΣΣbθfΣ(s)ds=ΣbθfΣ(s)dsηθzΣbθΨ(Σza)fΣ(a)dadμ(z). (19)

    Using the fact that the function Ψ is nonincreasing, we obtain the wanted inequality.

    Now we show that up to a subsequence, (fΣ)Σ converges to a fixed point of T denoted by f. Thanks to (18) and the properties of Ψ, we can define

    Σ0:=inf{Σ>max(11θb,1) such that ρΣ>12}.

    Lemma 3.7. Under hypotheses (2), (3), (4) and (5), the set of eigenfunctions {fΣ,ΣΣ0,fΣL1(R+,(sk+sl)ds)=1} has a compact closure in L1(R+,(sk+sl)ds), for any l0 and 0k<k0,k0 being the real number given in (5).

    Proof. Let l0 and k[0,k0). Once again, we apply Theorem 3.3 to show the desired result. First, we show that (14) hold true with Ω=(bθ,) and F={fΣ,fΣL1(R+,(sk+sl)ds)=1}. Let ω be a compact subset of (bθ,) and bθ<α<β such that ω[α,β]. We use the following inequality

    τhfΣfΣL1(ω,(sk+sl)ds)2τhTΣfΣTΣfΣL1(ω,(sk+sl)ds)2(βk+αl)τhTΣfΣTΣfΣL1([α,β])2(βk+αl)τhTΣfΣTΣfΣL1([α,Σ]).

    The last quantity is small when h is small uniformly with respect to Σ since in the proof of Lemma 3.4, the estimates do not depend on the value of Σ. To prove that (15) holds true, we use the estimate (16) twice to write

    fΣL1((bθ,α),(sk+sl)ds)=1ρΣαbθTΣfΣ(s)(sk+sl)ds2αbθTΣfΣ(s)slds+2αkαbθTΣfΣ(s)ds2θl(1Ψ(αbθθ))+2αk(1Ψ(αbθθ))2(θl+αk)(1Ψ(αbθθ))

    which is again independent of Σ. The estimate (17) though depends on Σ, so we write for Σ larger than β

    ρΣΣβfΣ(a)da=θΣβTΣfΣ(a)da+ηΣθΣTΣfΣ(a)da+ΣηΣTΣfΣ(a)da.

    For the first integral, we compute

    θΣβTΣfΣ(a)da=ηθzβzbθ[Ψ(βza)Ψ(θΣza)]fΣ(a)dadμ(z)+ηθzθΣzβz[1Ψ(θΣza)]fΣ(a)dadμ(z).

    The two other integrals correspond to the integrals (B), (C) and (D) from the previous proof. Combining the integrals, we obtain

    ρΣΣβfΣ(a)da=ηθzβzbθΨ(βza)fΣ(a)dadμ(z)+ηθzΣβzfΣ(a)dadμ(z)ηθzΣbθΨ(Σza)fΣ(a)dadμ(z).

    We deal with the last integral using (19) and obtain after interverting integrals

    ρΣΣβfΣ(a)da=βηbθfΣ(a)ηθzΨ(βza)dμ(z)da+βθβηfΣ(a)βaθzΨ(βza)dμ(z)da+βθβηfΣ(a)ηβazdμ(z)da+ΣβθfΣ(a)da+ρΣΣbθfΣ(a)daΣbθfΣ(a)daβηbθfΣ(a)da=βηbθfΣ(a)ηθzΨ(βza)dμ(z)da+βθβηfΣ(a)βaθzΨ(βza)dμ(z)da+ρΣβbθfΣ(a)daβθβηfΣ(a)βaθzdμ(z)daβηβfΣ(a)ηθz[1Ψ(βza)]dμ(z)da+(1ρΣ)βbθfΣ(a)da+βθβηfΣ(a)βaθz[1Ψ(βza)]dμ(z)da=βbθfΣ(a)ηθzΨ(βza)dμ(z)da

    Since 0<θ<η<1, we can choose β>η1ηbbθ. In that case, 1ηbβ>1, and we can pick r]1,1ηbβ[ such that (1ηr)β>b. Noticing that 1Ψ(βza) and 1ρΣ are nonnegative, we obtain

    rββfΣ(a)ηθz[1Ψ(βza)]dμ(z)daβbθfΣ(a)ηθzΨ(βza)dμ(z)da,

    then

    (1Ψ((1ηr)β))rββfΣ(a)daΨ((1η1)β),

    and finally

    rββfΣ(a)(al+ak)da(βl+(rβ)k)Ψ((1η1)β)(1Ψ((1ηr)β))4(rβ)kΨ((1η1)β)

    for β large enough. We use this estimate to get

    βfΣ(s)(sk+sl)ds=j=0rj+1βrjβfS(s)(sk+sl)ds4rkj=0(rjβ)kΨ((1η1)rjβ)
    4Crkj=0(rjβ)k((1η1)rjβ)k0Ck,k0,η,rβk0k

    due to hypothesis (5), for β large enough.

    We are now ready to prove the existence and uniqueness of a fixed point for the operator T.

    Proof of Proposition 2. We have proved in Lemma 3.7 that the set of eigenfunctions {fΣ, fΣL1(R+,(sl+sk)ds)} has a compact closure in L1(R+,(sk+sl)ds). We deduce the existence of fL1(R+,(sk+sl)ds) such that, up to a subsequence still denoted by (fΣ)Σ, fΣf strongly as Σ+. Now we prove that the function f is a fixed point of the operator T. We use the following inequality

    fTfL1(R+,(sk+sl)ds)ffΣL1(R+,(sk+sl)ds)+(1ρΣ)+TΣfΣTfL1(R+,(sk+sl)ds).

    The first term of the right-hand side tends to zero as Σ tends to by definition of f, and the second one is smaller than Ψ((1η1)Σ) according to (18). For the last one, we write

    TΣfΣTfL1(R+,(sk+sl)ds)TΣfΣTfΣL1(R+,(sk+sl)ds)=0+T(ffΣ)L1(R+,(sk+sl)ds)ffΣL1(R+,(sk+sl)ds)

    due to Lemma 3.2 and to the continuity of T, which is proved in Lemma 2.2.

    To prove uniqueness of the fixed point, we consider f1 another nonnegative fixed point of T in L1(R+) satisfying 0f1(s)ds=0f(s)ds. Recalling the definition (12) of the operator T, the functions f and f1 satisfy the integral convolution equation

    f(s)=ηθΦf(sz)dμ(z). (20)

    Since f, f1 and Φ are in L1(R+), their Laplace transforms exist on R+ and are continuous decreasing functions. Taking the Laplace transform of ff1 and switching integrals thanks to Fubini's theorem, one has for every y0

    L[ff1](y)=ηθL[ff1](zy)L[Φ](zy)zdμ(z). (21)

    The Laplace transform L[ff1] is continuous on R+ and vanishes at the origin

    L[ff1](0)=0f(s)ds0f1(s)ds=0.

    We now define the functions

    ¯L(y)=supx[0,y]L[ff1](x) and L_(y)=infx[0,y]L[ff1](x).

    By continuity in 0 of L[ff1] and because L[ff1](0)=0, one has

    y0,¯L(y)0, L_(y)0.

    From (21), we obtain the inequality

    L[ff1](y)¯L(ηy)ηθL[Φ](zy)zdμ(z)¯L(ηy),

    since Φ is a probability measure. ¯L is a continuous increasing function, so for all xy, one has

    L[ff1](x)¯L(ηx)¯L(ηy),

    from which we deduce

    ¯L(y)¯L(ηy). (22)

    Iterating (22), we obtain for all y0 and all positive integer j

    ¯L(y)¯L(ηjy).

    Letting j in this inequality and using the continuity of the function ¯L we obtain ¯L(y)=0 for all nonnegative y. With the same method, we show that L_(y)=0 for all nonnegative y, and finally L[ff1] is the null function. By the injectivity of the Laplace transform (Lerch's theorem [11]), one has f=f1.

    It remains to prove that suppf=[bθ,). With the same kind of proof than the one we used for TΣ, we can prove that T is irreducible on L1(bθ,), and since f is not the zero function we get the result.

    We are now ready to prove the main theorem of the paper.

    Proof of Theorem 1.1. Combining Lemma 2.1 and Proposition 2, we construct a solution to (11) using

    M(a,s):=ψ(a)(a+s)2f(s).

    It remains to prove uniqueness of the solution in the appropriate space. This solution belongs to L1(R2+,(1+s2)dads) thanks to the following calculation

    bθ0M(a,s)(1+s2)dads=bθ01(a+s)2f(s)Ψ(a)dads+bθ0s2(a+s)2f(s)Ψ(a)dadsbθf(s)s20Ψ(a)dads+bθf(s)0Ψ(a)dads=fL1((bθ,),(1+s2)ds)<

    because fL1((bθ,),slds) for all nonpositive number l. To prove the uniqueness of the solution ML1(R2+,(1+s2)dads), consider another solution M1L1(R2+,(1+s2)dads). Necessarily, as for M, there exists a measurable function f1 such that for almost all sa0

    M1(a,s)=Ψ(a)(a+s)2f1(s).

    For 0<α<β<, we can write

    0f1(s)ds=1βα0βα(a+s)2Ψ(a)M1(a,s)dads2(β2+1)(βα)Ψ(β)M1L1(R+,(1+s2)ds),

    and this ensures that f1L1(R+). Additionally we easily check as in Lemma 2.1 that f1 has to be a fixed point of T. Then the uniqueness result in Proposition 2 ensures that f1=f, and so M1=M. The existence and uniqueness of a solution to the initial problem (8) follows from the relation (10).

    Now that we have solved the eigenvalue problem, we would like to characterize the asymptotic behaviour of a solution n of (1), as in [18]. The General Relative Entropy principle provides informations about the evolution of the distance in L1 norm between a solution n(t,,) and etN. To establish such useful inequalities, we use the formalism introduced in [13] and [14]. Strictly speaking, to use this method, we should prove some properties on a time-dependent solution n, in particular its existence and uniqueness for any reasonable initial condition. Let us here assume the existence of such a solution, which moreover satisfies the common estimate (see [18])

    |n(t,a,x)|CetN(a,x),t0, x>a>0. (23)

    It is usually ensured by the hypothesis |n0(a,x)|CN(a,x) and a maximum principle. For H a function defined on all R, we define, for nL1(R2+)

    H[n]=bθxbθ0xN(a,x)H(n(a,x)N(a,x))dadx

    which satisfies the following entropy property.

    Proposition 3. If n is a solution of (1) satisfying (23), then

    ddtH[n(t,,)et]=D[n(t,,)et], (24)

    with

    D[n]=bθx2N(0,x)[ηθxzbθ0H(n(a,xz)N(a,xz))dνx(a,z)H(ηθxzbθ0n(a,xz)N(a,xz)dνx(a,z))]dx

    where dνx(a,z)=B(a)N(a,xz)N(0,x)z2dadμ(z) is a probability measure. Furthermore if H is convex, then D0.

    Before proving this proposition, we make a remark about the conservative problem (i.e. when only one daughter out of two is kept after division). In this case, the dominant eigenvalue is 0 instead of 1, and xN(a,x) is an eigenvector associated with the eigenvalue 0, since the total mass is preserved. Then we obtain the equation

    a(x2N)+x(x2N)=x2BN, (25)

    which might also be obtained multiplying (8a) by x.

    Proof. Easy computations lead to

    tnetN+xanetN+xxnetN=0,

    where N(a,x)>0, i.e. on the domain Ωθ:={xa>bθ}. From this equality and (25), we deduce

    t(xNH(netN))+a(x2NH(netN))+x(x2NH(netN))=x2BNH(netN), (26)

    and integrating (26) over Ωθ, we obtain

    ddt(bθ,)×(0,xbθ)xNH(netN)dadx=bθx2N(0,x)H(n(t,0,x)etN(0,x))dxbθx2N(xbθ,x)H(n(t,xbθ,x)etN(xbθ,x))dx+0(a+bθ)2N(a,a+bθ)H(n(t,a,a+bθ)etN(a,a+bθ))dabθxbθ0x2BNH(netN)dadx=bθx2N(0,x)H(etN(0,x)ηθxzbθ0B(a)n(t,a,xz)dadμ(z)z2)dx2ηθbθxbθ0x2BNH(netN)dadxzdμ(z)=bθx2N(0,x)H(ηθxzbθ0n(t,a,xz)etN(a,xz)dνx(a,z))dx2ηθbθzbθxzbθ0x2B(a)N(a,xz)H(n(t,a,xz)etN(a,xz))dadxdμ(z)z22ηθbθxzbθ0x2B(a)N(a,xz)H(n(t,a,xz)etN(a,xz))dadxdμ(z)z2=bθx2N(0,x)[H(ηθxzbθ0n(t,a,xz)etN(a,xz)dνx(a,z))ηθxzbθ0H(n(t,a,xz)etN(a,xz))dνx(a,z)]dx,

    since for x[zbθ,bθ] and z[θ,η],xzbθb, and we conclude using Jensen's inequality.

    Appropriate choices of the function H in (24) lead to interesting results. With H(x)=x, we recover the conservation law (7). Then taking H(x)=|1x|, we obtain the decay of NnetL1(Ωθ,xdxda) as t tends to infinity. In the case where the fragmentation kernel μ has a density with respect to the Lebesgue measure on [0,1], we expect that this quantity will vanish, as in [14,18]. In contrast, in the case of the equal mitosis, there is not hope for this distance to vanish. Indeed, one has an infinite number of eigentriplets (λj,Nj,ϕj) with jZ defined by

    λj=1+2ijπlog2,Nj(a,x)=x1λjN(a,x),ϕj(a,x)=xλj,

    so we expect a behavior as in [2], i.e. the convergence of n(t,a,x)et to the periodic solution

    jZn0,ϕje2ijπtlog2Nj(a,x),

    where n,ϕ=n(a,x)ϕ(a,x)dadx.

    We have proved the existence and uniqueness of a solution of the eigenproblem (8) in the special yet biologically relevant case of linear growth rate with a self-similar fragmentation kernel. Hypotheses on both this kernel and the division rate are fairly general.

    As possible future work we can imagine to extend the result to general growth rates. In this case the Perron eigenvalue is not explicit and it has to be determined in the same time as the eigenfunction, as in [26,12,6]. If we denote by λ the eigenvalue, the equivalent of Equation (20) is

    Pλ(s)=10eλszsdug(u)(ΦPλ)(sz)dμ(z)z

    with Pλ(y)=eλs1dug(u)M(0,s) and the equivalent of the solution given in (9) is

    N:(a,x)Ψ(a)g(x)eλx0dαg(α)Pλ(xa).

    Additionally for nonlinear growth rates, the function (a,x)x does not provide a conservation law as in (7), and it has to be replaced by a solution to the dual Perron eigenproblem. Such a dual eigenfunction appears in the definition of the General Relative Entropy [13,14], and for proving its existence one could follow the method in [19,8] for the size-structured model. Another possible generalization of the growth rate is adding variability, in the spirit of [21,15,17]. One might also consider a more general fragmentation kernel than in the case of self-similar fragmentation, or/and with a support which is not a compact subset of (0,1).

    The other natural continuation of the present work is the proof of the well-posedness and the long-time behavior of the evolution equation, as in [26,12]. To do so one can take advantage of the General Relative Entropy as in [14,4,2] or use general spectral methods [28,16].

    The authors are very grateful to Marie Doumic for having suggested them the problem treated in this paper, and for the many fruitful discussions.



    [1] V. I. Karpman, Stabilization of soliton instabilities by higher-order dispersion: fourth-order nonlinear Schrödinger-type equations, Phys. Rev. E, 53 (1996), R1336. https://doi.org/10.1103/physreve.53.r1336 doi: 10.1103/physreve.53.r1336
    [2] V. I. Karpman, A. G. Shagalov, Stability of solitons described by nonlinear Schrödinger-type equations with higher-order dispersion, Physica D, 144 (2000), 194–210. https://doi.org/10.1016/S0167-2789(00)00078-6 doi: 10.1016/S0167-2789(00)00078-6
    [3] Y. Fukumoto, H. K. Mofatt, Motion and expansion of a viscous vortex ring. Part 1. A higher-order asymptotic formula for the velocity, J. Fluid Mech., 417 (2000), 1–45. https://doi.org/10.1017/S0022112000008995 doi: 10.1017/S0022112000008995
    [4] T. Cazenave, P. L. Lions, Orbital stability of standing waves for some nonlinear Schrödinger equations, Commun. Math. Phys., 85 (1982), 549–561. https://doi.org/10.1007/BF01403504 doi: 10.1007/BF01403504
    [5] D. Bonheure, J. B. Casteras, T. Gou, L. Jeanjean, Normalized solutions to the mixed dispersion nonlinear Schrödinger equation in the mass critical and supercritical regime, Trans. Amer. Math. Soc., 372 (2019), 2167–2212. https://doi.org/10.1090/tran/7769 doi: 10.1090/tran/7769
    [6] O. Goubet, I. Manoubi, Standing waves for semilinear Schrödinger equations with discontinuous dispersion, Rend. Circ. Mat. Palermo, Ser. 2, 71 (2022), 1159–1171. https://doi.org/10.1007/s12215-022-00782-3 doi: 10.1007/s12215-022-00782-3
    [7] M. Khiddi, L. Essafi, Infinitely many solutions for quasilinear Schrödinger equations with sign-changing nonlinearity without the aid of 4-superlinear at infinity, Demonstr. Math., 55 (2022), 831–842. https://doi.org/10.1515/dema-2022-0169 doi: 10.1515/dema-2022-0169
    [8] M. Alotaibi, M. Jleli, B. Samet, C. Vetro, First and second critical exponents for an inhomogeneous Schrödinger equation with combined nonlinearities, Z. Angew. Math. Phys., 73 (2022), 157–173. https://doi.org/10.1007/s00033-022-01784-y doi: 10.1007/s00033-022-01784-y
    [9] D. Bonheure, J. B. Casteras, E. M. dos Santos, R. Nascimento, Orbitally stable standing waves of a mixed dispersion nonlinear Schrödinger equation, SIAM J. Math. Anal., 50 (2018), 5027–5071. https://doi.org/10.1137/17M1154138 doi: 10.1137/17M1154138
    [10] A. J. Fernández, L. Jeanjean, R. Mandel, M. Maris, Non-homogeneous Gagliardo-Nirenberg inequalities in RN and application to a biharmonic non-linear Schrödinger equation, J. Differ. Equations, 330 (2022), 1–65. https://doi.org/10.1016/j.jde.2022.04.037 doi: 10.1016/j.jde.2022.04.037
    [11] D. Bonheure, J. B. Casteras, T. Gou, L. Jeanjean, Strong instability of ground states to a fourth order Schrödinger equation, Int. Math. Res. Not., 2019 (2019), 5299–5315. https://doi.org/10.1093/imrn/rnx273 doi: 10.1093/imrn/rnx273
    [12] B. H. Feng, J. Y. Liu, H. L. Niu, B. L. Zhang, Strong instability of standing waves for a fourth-order nonlinear Schrödinger equation with the mixed dispersion, Nonlinear Anal., 196 (2020), 111791. https://doi.org/10.1016/j.na.2020.111791 doi: 10.1016/j.na.2020.111791
    [13] C. Ji, N. Su, Existence and stability of standing waves for the mixed dispersion nonlinear Schrödinger equation with a partial confinement in RN, J. Geom. Anal., 33 (2023), 171–195. https://doi.org/10.1007/s12220-023-01207-y doi: 10.1007/s12220-023-01207-y
    [14] H. J. Luo, Z. T. Zhang, Existence and stability of normalized solutions to the mixed dispersion nonlinear Schrödinger equations, Electron. Res. Arch., 30 (2022), 2871–2898. https://doi.org/10.3934/era.2022146 doi: 10.3934/era.2022146
    [15] Z. J. Ma, X. J. Chang, Normalized ground states of nonlinear biharmonic Schrödinger equations with Sobolev critical growth and combined nonlinearities, Appl. Math. Lett., 135 (2023), 108388. https://doi.org/10.1016/j.aml.2022.108388 doi: 10.1016/j.aml.2022.108388
    [16] E. D. Silva, M. L. Carvalho, C. Goulart, Periodic and asymptotically periodic fourth-order Schrödinger equations with critical and subcritical growth, Discrete Contin. Dyn. Syst., 42 (2022), 1039–1065. https://doi.org/10.3934/dcds.2021146 doi: 10.3934/dcds.2021146
    [17] C. O. Alves, C. Ji, O. H. Miyagaki, Multiplicity of normalized solutions for a nonlinear Schrödinger equation with critical growth in RN, preprint, arXiv: 2103.07940.
    [18] C. O. Alves, C. Ji, O. H. Miyagaki, Normalized solutions for a Schrödinger equation with critical growth in RN, Calc. Var. Partial Differ. Equations, 61 (2022), 18–41. https://doi.org/10.1007/s00526-021-02123-1 doi: 10.1007/s00526-021-02123-1
    [19] H. Berestycki, P. L. Lions, Nonlinear scalar field equations, I existence of a ground state, Arch. Ration. Mech. Anal., 82 (1983), 313–345. https://doi.org/10.1007/BF00250555 doi: 10.1007/BF00250555
    [20] M. J. Esteban, P. L. Lions, Stationary solutions of nonlinear Schrödinger equations with an external magnetic field, in Partial Differential Equations and the Calculus of Variations, (1989), 401–449. https://doi.org/10.1007/978-1-4615-9828-2_18
    [21] P. L. Lions, The concentration-compactness principle in the calculus of variations, The locally compact case, Part I, Ann. Inst. Henri Poincare C, 1 (1984), 109–145. https://doi.org/10.1016/S0294-1449(16)30428-0 doi: 10.1016/S0294-1449(16)30428-0
    [22] C. O. Alves, C. Ji, Normalized solutions for the Schrödinger equations with L2-subcritical growth and different types of potentials, J. Geom. Anal., 32 (2022), 165–189. https://doi.org/10.1007/s12220-022-00908-0 doi: 10.1007/s12220-022-00908-0
    [23] L. Jeanjean, Existence of solutions with prescribed norm for semilinear elliptic equations, Nonlinear Anal. Theory Methods Appl., 28 (1997), 1633–1659. https://doi.org/10.1016/S0362-546X(96)00021-1 doi: 10.1016/S0362-546X(96)00021-1
    [24] M. Shibata, Stable standing waves of nonlinear Schrödinger equations with a general nonlinear term, Manuscr. Math., 143 (2014), 221–237. https://doi.org/10.1007/s00229-013-0627-9 doi: 10.1007/s00229-013-0627-9
    [25] E. Gagliardo, Caratterizzazioni delle tracce sulla frontiera relative ad alcune classi di funzioni in n variabili, Rend. Semin. Mat. Univ. di Padova, 27 (1957), 284–305.
    [26] E. Gagliardo, Proprietà di alcune classi di funzioni in più variabili, Ricerche Mat., 7 (1958), 102–137. Available from: https://zbmath.org/?q = an: 0089.09401.
    [27] L. Nirenberg, On elliptic partial differential equations, Ann. Sc. Norm. Super. Pisa. Cl. Sci., 13 (1959), 115–162. Available from: http://www.numdam.org/item/ASNSP_1959_3_13_2_115_0.pdf.
    [28] M. Willem, Minimax Theorems, 1996. https://doi.org/10.1007/978-1-4612-4146-1
    [29] T. Bartsch, Z. Q. Wang, Multiple positive solutions for a nonlinear Schrödinger equation, Z. Angew. Math. Phys., 51 (2000), 366–384. https://doi.org/10.1007/PL00001511 doi: 10.1007/PL00001511
  • This article has been cited by:

    1. Bertrand Cloez, Benoîte de Saporta, Tristan Roget, Long-time behavior and Darwinian optimality for an asymmetric size-structured branching process, 2021, 83, 0303-6812, 10.1007/s00285-021-01695-y
    2. Pierre Gabriel, Hugo Martin, Periodic asymptotic dynamics of the measure solutions to an equal mitosis equation, 2022, 5, 2644-9463, 275, 10.5802/ahl.123
    3. Katarzyna Pichór, Ryszard Rudnicki, Cell cycle length and long‐time behavior of an age‐size model, 2022, 45, 0170-4214, 5797, 10.1002/mma.8139
    4. Mingtao Xia, Chris D. Greenman, Tom Chou, PDE Models of Adder Mechanisms in Cellular Proliferation, 2020, 80, 0036-1399, 1307, 10.1137/19M1246754
    5. Ignacio Madrid, Exponential Ergodicity of a Degenerate Age-Size Piecewise Deterministic Process, 2023, 187, 0167-8019, 10.1007/s10440-023-00597-z
  • Reader Comments
  • © 2023 the Author(s), licensee AIMS Press. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0)
通讯作者: 陈斌, bchen63@163.com
  • 1. 

    沈阳化工大学材料科学与工程学院 沈阳 110142

  1. 本站搜索
  2. 百度学术搜索
  3. 万方数据库搜索
  4. CNKI搜索

Metrics

Article views(1570) PDF downloads(128) Cited by(2)

Other Articles By Authors

/

DownLoad:  Full-Size Img  PowerPoint
Return
Return

Catalog