\documentclass[reqno]{amsart}
\usepackage{hyperref}

\AtBeginDocument{{\noindent\small
\emph{Electronic Journal of Differential Equations},
Vol. 2016 (2016), No. 86, pp. 1--15.\newline
ISSN: 1072-6691. URL: http://ejde.math.txstate.edu or http://ejde.math.unt.edu
\newline ftp ejde.math.txstate.edu}
\thanks{\copyright 2016 Texas State University.}
\vspace{9mm}}

\begin{document}
\title[\hfilneg EJDE-2016/86\hfil Exponential stability of traveling waves]
{Exponential stability of traveling waves for non-monotone
delayed reaction-diffusion equations}

\author[Y. Liu, Z. Yu,  J. Xia \hfil EJDE-2016/86\hfilneg]
{Yixin Liu, Zhixian Yu, Jing Xia}

\address{Yixin Liu \newline
College of Science,
University of Shanghai for Science and Technology,
Shanghai, 200093, China}
\email{838080693@qq.com}

\address{Zhixian Yu (corresponding author)\newline
College of Science,
University of Shanghai for Science and Technology,
Shanghai, 200093, China}
\email{zxyu@usst.edu.cn, zxyu0902@163.com}


\address{Jing Xia \newline
Department of Fundamental Courses,
Academy of Armored Force Engineering,
Beijing 100072, China}
\email{xiajingcjh@163.com}

\thanks{Submitted July 20, 2015. Published March 29, 2016.}
\subjclass[2010]{35C07, 92D25, 35B35}
\keywords{Stability; non-monotone; weighted energy method}

\begin{abstract}
 This article concerns the  exponential stability of non-critical
 traveling waves (the wave speed is greater than the minimum speed)
 for non-monotone time-delayed reaction-diffusion equations.
 With the help of the weighted energy method, we prove that the non-critical
 travelling waves are exponentially stable when the initial perturbation
 around the wave is small.
\end{abstract}

\maketitle
\numberwithin{equation}{section}
\newtheorem{theorem}{Theorem}[section]
\newtheorem{lemma}[theorem]{Lemma}
\newtheorem{proposition}[theorem]{Proposition}
\allowdisplaybreaks

\section{Introduction}

 In this article, we study the stability of traveling waves for
 the non-monotone delayed reaction diffusion equation
\begin{equation}\label{1.1}
\frac{\partial\upsilon(t,x)}{\partial t}
=D\frac{\partial^2\upsilon(t,x)}{\partial x^2}-d(\upsilon(t,x))+f(\upsilon(t-r,x)),
 \quad( t,x) \in \mathbb{R}^+\times\mathbb{R}
\end{equation}
with the initial condition
\begin{equation}\label{1.2}
\upsilon(s,x)=\upsilon_0(s,x),\quad s\in[-r,0],\; x\in\mathbb{R}.
\end{equation}
where $D>0$, $r\geq 0$ are constants. The nonlinear functions
$d(u)$ and $f(u)$ satisfy the following hypotheses:
 \begin{itemize}
  \item [(H1)] $d\in C^2([0,\infty],\mathbb{R})$,
$f \in C^2 ([0,\infty],\mathbb{R})$; there exist only two constant equilibria
$0$ and $K>0$ such that $f(0)=d(0)$, $f(K)=d(K)$, $d'(0)-f'(0)<0$ and
$d'(K)-f'(K)>0$.

  \item [(H2)] There exists $K^\ast\geq K$ such that
$d(K^*)\geq \max\{f(\upsilon)|0\leq \upsilon\leq K^*\}$ and
$d(\upsilon)<d(K^*)$ for all $\upsilon\in [0,K^*)$,
$f'(0)\upsilon\geq f(\upsilon)>0$, $d(\upsilon)\geq d'(0)\upsilon$ and
$f'(0)\upsilon>d(\upsilon)$ for all $\upsilon \in (0,K^\ast]$.

\item [(H3)]  $d(\upsilon)$ is strictly increasing on $[0,K^\ast]$ and
$d(\upsilon)<f(\upsilon)<2d(K)-d(\upsilon)$ for $\upsilon \in[0,K)$,
$d(\upsilon)>f(\upsilon)>2d(K)-d(\upsilon)$ for $\upsilon \in (K,K^\ast]$.

\item [(H4)] $d'(\upsilon)\geq d'(0)$ and $|f'(\upsilon)|\leq f'(0)$
for all $\upsilon \in (0,K^\ast]$.
\end{itemize}


Equation \eqref{1.1} includes several practical models.
Letting $d(\upsilon(t,x))=\delta\upsilon(t,x)$, \eqref{1.1} reduces to the
 time-delayed reaction-diffusion equation
\begin{equation}\label{1.3}
\frac{\partial\upsilon(t,x)}{\partial t}
=D\frac{\partial^2\upsilon(t,x)}{\partial x^2}-\delta\upsilon(t,x)
+f(\upsilon(t-r,x)).
\end{equation}
This model represents the single species population distribution such as the
Australian blowfly \cite{Gourley-So-Wu,Gourley-Wu,So-Yang}.
 Here $\upsilon(t,x)$ denotes the mature population of the blowflies
at location $ x$ and time $t$, $ D>0$ and $\delta> 0$ are the diffusion coefficient
and death rate of the mature population, the time delay $r>0$ is the time
taken from birth to maturity, and $f(\upsilon(t-r,x))$ is the birth function.
Especially, taking $f(\upsilon)=p \upsilon e^{-a \upsilon}$, $p>0$, $a>0$,
\eqref{1.3} is a typical Nicholson's blowflies model; i.e.,
\begin{equation}\label{1.4}
\frac{\partial\upsilon(t,x)}{\partial t}
=D\frac{\partial^2\upsilon(t,x)}{\partial x^2}
-\delta\upsilon(t,x)+p \upsilon(t-r) e^{-a \upsilon(t-r)}.
\end{equation}

When the birth function $f$ is monotone, authors in
\cite{Gomez-Trofimchuk,Lin-Mei,Mei1,Mei2,So-Zou,Trofimchuk-Pinto}
investigated the existence of monotone traveling waves by using the
 monotone iteration and fixed-points theorem with help of the upper-lower
solutions. Schaaf \cite{Schaaf} first studied linear stability for the
delayed reaction diffusion with the quasi-monotone nonlinear terms,
which includes \eqref{1.1}, by using a spectral method.
The authors in \cite{Mei-So-Li}  investigated the nonlinear stability of
traveling waves by using the (technical) weighted energy method.
Then authors in \cite{Mei-Wang} further employed its global stability
 by using the weighted energy technique and the comparison principle.
These results were then extended to more general delayed reaction diffusion
equations with the quasi-monotone nonlinearity in
 \cite{Lv-Wang,Lv-Wang1,Wu-Li-Liu}. By using the Fourier transform,
Green's function and the weighted energy method,
the authors in \cite{Mei-Ou-Zhao,Mei-Wang} showed the global stability
 of critical traveling waves, which depends on the monotonicity
of both the equation and traveling waves.

However, because of the lack of monotonicity, the simple but useful methods
have failed. For this case
\cite{Faria-Huang-Wu,Faria-Trofimchuk,Faria-Trofimchuk1,Ma,Wu-Li-Liu,
Gomez-Trofimchuk1} show the existence of traveling waves by developing
different methods. Especially, the study on the stability of traveling
waves is quite limited. Wu, Zhao and Liu \cite{Wu-Zhao-Liu} first showed
the stability of traveling waves with the large wave speed for \eqref{1.3}
by using weighted energy method. Recently, Lin et al \cite{Chi-Mei}
established the stability of traveling waves (including oscillating
traveling waves) for \eqref{1.4} by using the weighted energy method and
the nonlinear Halanay inequality. Then Chern et al \cite{Chern-Mei}
followed the recent study \cite{Chi-Mei} and further answered all critical
traveling waves for \eqref{1.4} are time-asymptotically stable with the help
of some new development.

To the best of our knowledge, the stability of traveling waves for the more
general non-monotone delayed reaction diffusion equation \eqref{1.1}
is still not investigated. The methods in \cite{Wu-Zhao-Liu,Chern-Mei}
can still be used owing to the boundedness of the solution with the initial
condition for the non-monotone delayed reaction diffusion equation \eqref{1.1},
 which was proved in \cite{Wang-Li}.


\section{Preliminaries and main result}

We first introduce some notation. Throughout this paper, $C>0$ denotes a
generic constant, while $C_i>0$ $(i=0,1,2\dots)$ represents a special constant.
Letting $I$ be an interval, especially $I=\mathbb{R}$, $L^2(I)$
is the space of the square integrable function on $I$, and $H^k(I)(k\geq 0)$
is the Sobolev space of the $L^2$-function  $f(x)$
defined on $I$ whose derivatives $\frac {d^i}{dx^i}f$, $i=1,\dots,k$, also
belong to $L^2(I)$. $L_\omega ^2(I)$ represents the weighted
$L^2$-space with the weight $\omega(x)>0$ and its norm is defined by
$$
\|f\|_{L_\omega^2}=\Big(\int_{I} \omega(x)f^2(x)dx\Big)^{1/2}.
$$
$H_\omega^k(I)$ is the weighted Sobolev space with the norm given by
$$
\|f\|_{H_\omega^k}=\Big(\sum_{{i=0}}^{k}\int_{I} \omega(x)|
\frac{d^i}{dx^i}f(x)|^2 dx\Big)^{1/2}.
$$
Letting $T>0$ and $B$ space, we denote by $C^0([0,T];\textit{B})$ the space
of the $B$-valued continuous
functions on $[0,T]$, and $L^2 ([0,T];\textit{B})$ as the space of $B$-valued
$L^2$-function on $[0,T]$. The corresponding spaces
of the $B$-valued function on $[0,\infty)$ are defined similarly.

The traveling waves for \eqref{1.1} connecting 0 and $K$ are the special
 solution to \eqref{1.1} in the form of $\upsilon(t,x)=\phi(x+ct)$,
namely, $\phi$ satisfies
\begin{gather}\label{1.40}
c\phi'(\xi)-D\phi''(\xi)+d(\phi(\xi))-f(\phi(\xi-cr))=0, \\
\label{1.5}
\phi(-\infty)=0,\quad \phi(+\infty)=K.
\end{gather}

If $f'(0)>d'(0)$, there exists a unique number $c_*>0$ such that for $c>c_*$,
the characteristic equation $\Delta(c,\lambda)=0$ of linearized equation
at 0 for \eqref{1.40} has two positive roots $\lambda_1=\lambda_1(c)>0$ and
$\lambda_2=\lambda_2(c)>0$, where
\begin{equation}\label{1.6}
\Delta(c,\lambda):=c\lambda-D\lambda^2+d'(0)-e^{-\lambda cr}f'(0).
\end{equation}
Moreover,
\begin{equation}\label{1.7}
c\lambda-D\lambda^2+d'(0)>e^{-\lambda cr}f'(0),\quad \text{for }
 \lambda_1<\lambda<\lambda_2.
\end{equation}

Let us recall the existence and uniqueness of traveling waves for \eqref{1.1}
with the non-monotone nonlinearity, (see \cite{Ma}) and some related
results can also be found in \cite{Fang-Zhao,Wu-Liu} and the boundedness
of the solution with the initial condition for the non-monotone delayed
reaction diffusion equation \eqref{1.1}, see  \cite{Wang-Li}.

\begin{proposition}\label{prop2.1}
Assume that {\rm (H1)--(H3)} hold, there exists a unique number
$c_*>0$ such for every $c>c_\ast$, Equation \eqref{1.1} has a unique
(up to translation) traveling wave solution $\phi(\xi)$ satisfying
$\phi(-\infty)=0$, $\phi(+\infty)=K$ and $0\leq\phi(\xi)\leq K^\ast$ for all
$\xi\in \mathbb{R}$.
\end{proposition}


\begin{proposition}\label{prop2.2}
 Assume that {\rm (H1)--(H4)} hold and
$0\leq\upsilon_0(s,x)\leq K^\ast$ for all $(s,x)\in [-r,0]\times\mathbb{R}$.
Then the solution of Cauchy problem \eqref{1.1} and \eqref{1.2} satisfies
$$
0\leq\upsilon(t,x)\leq K^\ast   \quad\text{for all }\
 (t,x)\in [0,+\infty)\times\mathbb{R}.
$$
\end{proposition}

For $\lambda_1<\lambda<\lambda_2$ and some number $\xi_\ast$, define the
weight function
\begin{equation}\label{1.9}
\omega(\xi)=\begin{cases}
e^{-2\lambda(\xi-\xi_\ast)},&\text{for }\xi<\xi_\ast,\\
1,                          &\text{for } \xi\geq\xi_\ast.
\end{cases}
\end{equation}
For a given weight function $\omega(\xi)$ and letting $T\geq0$, we define
the solution spaces as
$$
X(-r,T)=\{u|u(t,\xi)\in C([-r,T];C(\mathbb{R})\cap H_\omega^1(\mathbb{R}))\}
$$
and
$$
M(T)^2=\sup_{t\in[-r,T]}\Big(\|u(t)\|^2_{C}+\|u(t)\|^2_{H_\omega^1}\Big).
$$
In particular, when $T=\infty$, we can also define the solution space as
$X(-r,\infty)$ and the norm of the solution space as $M(\infty)$.
Now, we state the stability result for  \eqref{1.1}.

\begin{theorem}[Stability] \label{thm2.1}
 Assume that {\rm (H1)--(H4)} hold and $|f'(K)|$ is sufficiently small.
For any given traveling wave $\phi(\xi)$ of \eqref{1.4} with speed $c>c_*$,
if the initial perturbation is small; i.e.,
$$
\max_{s\in[-r,0]}\|(\upsilon_0-\phi)(s)\|^2_C
+\|(\upsilon_0-\phi)(0)\|^2_{H_\omega^1}
+\int_{-r}^0\|(\upsilon_0-\phi)(s)\|^2_{H_\omega^1}ds\leq\delta_0^2,
$$
then the unique solution $\upsilon(t,x)$ of \eqref{1.1} and \eqref{1.2}
exists globally and satisfies
\begin{gather}\label{3.5}
\upsilon(t,x)-\phi(x+ct)\in C([-r,\infty); C(\mathbb{R})\cap H_\omega^1(\mathbb{R})),\\
\label{3.6}
\sup_{x\in \mathbb{R}}|\upsilon(t,x)-\phi(x+ct)|\leq Ce^{-\mu t},\,\,\mbox\,\,t>0
\end{gather}
for the constant $C>0$ and $\mu>0$.
\end{theorem}

\section{Proof of stability}

To obtain the stability of non-monotone delayed reaction diffusion equations,
 we need to give some results.

\begin{lemma}\label{lem3.1}
Assume that {\rm (H1)--(H4)} and $\phi$ is a traveling wave for \eqref{1.4}.
Then there exists $A>0$ such that
\[
|\phi'(\xi)|\leq A  \quad \text{and} \quad
\lim_{\xi\to\pm\infty}\phi'(\xi)=0
\]
\end{lemma}

\begin{proof} Letting
\[
\rho_1=\frac{c-\sqrt{c^2+4Dd'(0)}}{2D}  \quad \text{and} \quad
\rho_2=\frac{c+\sqrt{c^2+4Dd'(0)}}{2D},
\]
it follows from \eqref{1.4} that
\begin{align*}
\phi(\xi)=\frac{1}{D(\rho_2-\rho_1)}
\Big[\int^{\xi}_{-\infty}e^{\rho_1(\xi-s)}H(\phi)(s)ds
+\int^{+\infty}_{\xi}e^{\rho_2(\xi-s)}H(\phi)(s)ds\Big],
\end{align*}
where $H(\phi)(s)=f(\phi(s-cr))+d'(0)\phi(\xi)-d(\phi(\xi))$.
Differentiating the above equation with respect to $\xi$, we obtain
\begin{equation}\label{2.1}
\phi'(\xi)=\frac{1}{D(\rho_2-\rho_1)}\Big[\int^{\xi}_{-\infty}
\rho_1e^{\rho_1(\xi-s)}H(\phi)(s)ds
+\int^{+\infty}_{\xi}\rho_2e^{\rho_2(\xi-s)}H(\phi)(s)ds\Big].
\end{equation}
Since $\rho_2-\rho_1\geq 2\sqrt{\frac{d'(0)}{D}}$, we obtain
\[
|\phi'(\xi)|\leq \frac{1}{\sqrt{Dd'(0)}}
\max_{s\in  \mathbb{R}}|H(\phi)(s)|:=A \quad \text{for all }
 \xi \in \mathbb{R}.
\]
Finally, \eqref{2.1} and the L.Hopital's rule imply that
$\lim_{\xi\to+\infty}\phi'(\xi)=0$.  the proof is complete.
\end{proof}

\begin{lemma}\label{lem2.2}
Assume that $f'(0)>d'(0)$ and $|f'(K)|$ is sufficiently small. Then,
for every $c>c_\ast$, there exist $\xi_0, \xi_\ast\in \mathbb{R}$ with
$\xi_\ast>\xi_0$ such that
\[
\max\{|f'(\phi(\xi_0-cr))|,|f'(\phi(\xi_0))|\}
\leq\frac{\min\{\Delta(c,\lambda)+e^{-\lambda cr}f'(0),d'(0)\}}{\cosh(\lambda cr)},
\]
and for $\xi\geq\xi_\ast-cr>\xi_0$,
\[
\max\{|f'(\phi(\xi-cr))|,|f'(\phi(\xi))|\}
\leq\max\{|f'(\phi(\xi_0-cr))|,|f'(\phi(\xi_0))|\}.
\]
\end{lemma}

 Since $\lim_{\xi\to\pm\infty}\phi(\xi)=K$ and $f'(K)$ is sufficiently small,
the conclusion of the above lemma obviously holds.

Letting $u(t,\xi):=\upsilon(t,x)-\phi(\xi), \xi=x+ct$, where $\phi(x+ct)$
is a given traveling wave solution of \eqref{1.1}, the Cauchy problem
\eqref{1.1} and \eqref{1.2} can be reformulated as
\begin{equation}\label{3.1}
\begin{gathered}
u_t(t,\xi)+cu_\xi(t,\xi)-Du_{\xi\xi}(t,\xi)
=g(u(t-r,\xi-cr))-p(u(t,\xi)),\\
 (t,\xi)\in (0,+\infty)\times \mathbb{R},\\
u(s,\xi)=\upsilon_0(s,\xi-cs )-\phi(\xi)=:u_0(s,\xi),\quad
(s,\xi)\in [-r,0]\times \mathbb{R},
\end{gathered}
\end{equation}
where
$$
g(u)=f(\phi+u)-f(\phi),\quad p(u)=d(\phi+u)-d(\phi).
$$
By the iteration technique and the energy method
(see \cite{Chi-Mei,Matsumura-Nishida,Mei} ), we can obtain the existence of
local solutions for \eqref{3.1}.

\begin{theorem}\label{thm3.1}
 Assume that {\rm (H1)--(H4)} hold.
For any given traveling wave $\phi(\xi)$ with $c>c_*$, suppose
$u_0(s,\xi)\in X(-r,0)$, and $M(0)\leq\delta_1$, where $\delta_1$
is a given positive constant. Then there exists a small $t_0=t_0(\delta_1)>0$
such that the local solution $u(t,\xi)$ of \eqref{3.1} uniquely exists for
$t\in[-r,t_0]$ and satisfies $u\in X(-r,t_0)$ and $M(t_0)\leq a M(0)$
for some constant $a$.
\end{theorem}

\begin{proof}
Let $u^{(0)}(t,\xi):=u_0(t,\xi)\in X(-r,0)\subseteq X(-r,t_0)$.
Then define the iteration $u^{(n+1)}=\mathcal{T}(u^{(n)})$ for $n\geq0$ by
\begin{equation}\label{3.4}
\begin{gathered}
\frac{\partial u^{(n+1)}}{\partial t}+c\frac{\partial u^{(n+1)}}{\partial \xi}
-D\frac{\partial^2 u^{(n+1)}}{\partial\xi^2}=g(u^{(n)}(t-r,\xi-cr))-p(u^{(n)}
 (t,\xi)),\\
u^{(n+1)}(s,\xi)=u_0(s,\xi),\quad s\in[-r,0],\; \xi\in \mathbb{R}.
\end{gathered}
\end{equation}
Using Fourier transform, \eqref{3.4} can be written as
\begin{equation}\label{3.8}
\begin{aligned}
u^{(n+1)}(t,\xi)
&=\int_\mathbb{R}\Gamma(\eta,t)u_0(0,\xi-\eta)d\eta
 +\int_0^t\int_\mathbb{R}\Gamma(\eta,t-s)\\
&\quad\times \big[g(u^{(n)}(s-r,\xi-\eta+cr))
  -p(u^{(n)}(s,\xi-\eta))\big]\,d\eta\,d\xi,
\end{aligned}
\end{equation}
where $\Gamma(\eta,t)$ is the heat kernel
$$
\Gamma(\eta,t)=\frac{1}{\sqrt{4\pi Dt}}e^{-\frac{(\eta+ct)^2}{4Dt}}.
$$
By applying regular energy estimates to both sides of \eqref{3.4}, 
indicated as 
$$
\int_0^t\int_\mathbb{R}\Big(\sum_{k=0}^1\partial_\xi^k(\eqref{3.4})
\times\omega(\xi)\partial_\xi^ku^{(n+1)}\Big)\,d\xi\,ds,
$$
we can estimate
\begin{equation}\label{3.9}
\begin{aligned}
&\|u^{(n+1)}(t)\|_{{H_\omega^1}}^2
 +\int_{0}^t \|u^{(n+1)}(s)\|_{{H_\omega^1}}^2 ds \\
&\leq C\Big(\|u_0(0)\|_{H^1_{\omega}}^2+\int_{-r}^0 \|u_0(s)\|_{{H_\omega^1}}^2  ds
+\int_{0}^t \|u^{(n)}(s)\|_{{H_\omega^1}}^2 ds\Big),\quad t\in[0,t_0]
\end{aligned}
\end{equation}
for some positive constant $C>0$. From \eqref{3.8} it follows that
\begin{equation}\label{3.10}
\|u^{(n+1)}(t)\|_C\leq C\|u_0(0)\|_C+Ct_0\sup_{t\in[-r,t_0]} \|u^{(n)}(t)\|_C,
\quad t\in[0,t_0].
\end{equation}
According to \eqref{3.9} and \eqref{3.10}, it holds that
\[
M_{u^{(n+1)}}(t_0)
\leq C\Big(\max_{s\in[-r,0]}\|u_0(s)\|^2_C+\|u_0(0)\|^2_{H_\omega^1}
+\int_{-r}^0\|u_0(s)\|^2_{H_\omega^1}ds\Big)
+Ct_0M_{u^{(n)}}(t_0).
\]
Thus,  when $\max_{s\in[-r,0]}\|u_0(s)\|^2_C+\|u_0(0)\|^2_{H_\omega^1}
+\int_{-r}^0\|u_0(s)\|^2_{H_\omega^1}ds\ll 1$ with
$0<t_0\ll 1$, $u^{(n+1)}=\mathcal{T}(u^{(n)})$ defined in \eqref{3.4}
is a contraction mapping from $X(-r,t_0)$ to $X(-r,t_0)$. Hence, by using
Banach fixed point theorem, \eqref{3.1} admits a unique local solution
in $X(-r,t_0)$. This completes the proof.
\end{proof}

\subsection*{A priori estimate}
 We rewrite \eqref{3.1} as
\begin{equation}\label{3.01}
\begin{gathered}
\begin{aligned}
&u_t(t,\xi)+cu_\xi(t,\xi)-Du_{\xi\xi}(t,\xi)+d'(\phi(\xi))u(t,\xi)
-f'(\phi(\xi-cr))u(t-r,\xi-cr)\\
&=G(u)(t,\xi)-E(u)(t,\xi),\quad (t,\xi)\in (0,+\infty)\times \mathbb{R},
\end{aligned}\\
u(s,\xi)=u_0(s,\xi),\quad (s,\xi)\in [-r,0]\times \mathbb{R},
\end{gathered}
\end{equation}
where
\begin{gather}\label{3.02}
\begin{aligned}
G(u)(t,\xi)&=f(u(t-r,\xi-cr)+\phi(\xi-cr))-f(\phi(\xi-cr))\\
&\quad -f'(\phi(\xi-cr))u(t-r,\xi-cr),
\end{aligned} \\
\label{3.03}
E(u)(t,\xi)=d(u(t,\xi)+\phi(\xi))-d(\phi(\xi))-d'(\phi(\xi))u(t,\xi).
\end{gather}

\begin{lemma}\label{lem4.1}
Let $u(t,\xi)\in X (-r,T)$. Then
\begin{equation}\label{4.1}
\begin{aligned}
&\|u(t)\|_{{L_\omega^2}}^2+\int_{0}^te^{-2\mu (t-s)}
\int_\mathbb{R} [B_{\eta,\mu,\omega}(\xi)-CM(t)]\omega(\xi)u^2(s,\xi)\,d\xi\,ds  \\
&\leq Ce^{-2\mu t}\Big(\|u_0(0)\|_{L^2_{\omega}}^2
 +\int_{-r}^0 \|u_0(s)\|_{{L_\omega^2}}^2 ds \Big),
\end{aligned}
\end{equation}
where
\begin{gather}\label{4.2}
B_{\eta,\mu,\omega}(\xi):=A_{\eta,\omega}(\xi)-2\mu
-\frac{1}{\eta}(e^{2\mu r}-1)|f'(\phi(\xi))|\frac{\omega(\xi+cr)} {\omega(\xi)},\\
\label{4.3}
\begin{aligned}
A_{\eta,\omega}(\xi)
:=&-c\frac{\omega'(\xi)} {\omega(\xi)}+2d'(\phi(\xi))
-\frac{D}{2}(\frac{ \omega'(\xi)}{\omega(\xi)})^2-\eta |f'(\phi(\xi-cr))|\\
&-\frac{1}{\eta}|f'(\phi(\xi))|\frac{\omega(\xi+cr)} {\omega(\xi)},
\end{aligned}
\end{gather}
and $\mu,\eta$ are positive constants.
\end{lemma}

\begin{proof}
 Multiplying \eqref{3.01} by $e^{2\mu t}\omega(\xi)u(t,\xi)$
with $ \xi\in \mathbb{R} $ and $ 0\leq t \leq T $, we have
\begin{equation}\label{4.4}
\begin{aligned}
&\Big\{\frac{1}{2}e^{2\mu t}\omega u^2\Big\}_t+e^{2\mu t}
\Big\{\frac{1}{2}c\omega u^2-D\omega u u_\xi\Big\}_\xi
+De^{2\mu t}\omega u_{\xi}^2+De^{2\mu t}\omega'u_{\xi}u \\
&+\Big\{-\frac{c}{2}\frac{\omega'}{\omega}+d'(\phi(\xi))-\mu\Big\}
e^{2\mu t}\omega u^2-e^{2\mu t}\omega(\xi) u(t,\xi) f'(\phi(\xi-cr))u(t-r,\xi-cr) \\
&=e^{2\mu t}\omega(\xi) u(t,\xi)[G(u)-E(u)].
\end{aligned}
\end{equation}
By the Cauchy-Schwarz inequality,
$$
|De^{2\mu t}\omega'u_{\xi}u|
\leq De^{2\mu t}\omega u_{\xi}^2+\frac {D}{4}e^{2\mu t}
\Big(\frac{\omega'}{\omega}\Big)^2\omega u^2,
$$
we reduce \eqref{4.4} to
\begin{equation}\label{4.5}
\begin{aligned}
&\Big\{\frac{1}{2}e^{2\mu t}\omega u^2\Big\}_t+e^{2\mu t}
\Big\{\frac{1}{2}c\omega u^2-D\omega u u_\xi\Big\}_\xi \\
&+\Big\{-\frac{c}{2}\frac{\omega'}{\omega}+d'(\phi(\xi))-\mu
-\frac {D}{4}\Big(\frac{\omega'}{\omega}\Big)^2\Big\}e^{2\mu t}\omega u^2 \\
&-e^{2\mu t}\omega(\xi) u(t,\xi) f'(\phi(\xi-cr))u(t-r,\xi-cr) \\
&\leq e^{2\mu t}\omega(\xi) u(t,\xi)[G(u)-E(u)].
\end{aligned}
\end{equation}
Integrating \eqref{4.5} over $\mathbb{R}\times[0,t]$
with respect to $\xi$ and $ t $, we have
\begin{equation}\label{4.6}
\begin{aligned}
&e^{2\mu t}\|u(t)\|_{{L_\omega^2}}^2+\int_{0}^t
\int_\mathbb{R} e^{2\mu s} \Big\{-c\frac{\omega'(\xi)}{\omega(\xi)}
+2d'(\phi(\xi))-2\mu-\frac {D}{2}\Big(\frac{\omega'(\xi)}
{\omega(\xi)}\Big)^2\Big\} \\
&\times \omega(\xi)u^2(s,\xi)\,d\xi\,ds  \\
&-2\int_{0}^t \int_\mathbb{R} e^{2\mu s}\omega(\xi)f'(\phi(\xi-cr))u(s,\xi)
u(s-r,\xi-cr)\,d\xi\,ds  \\
&\leq \|u_0(0)\|_{{L_\omega^2}}^2
+2\int_{0}^t\int_\mathbb{R} e^{2\mu s} \omega(\xi) u(s,\xi)
 [G(u)(s,\xi)-E(u)(s,\xi)]\,d\xi\,ds.
\end{aligned}
\end{equation}
Since
\begin{equation}\label{4.7}
\begin{aligned}
&2\Big|\int_{0}^t \int_\mathbb{R} e^{2\mu s}\omega(\xi)f'(\phi(\xi-cr))u(s,\xi)
  u(s-r,\xi-cr)\,d\xi\,ds\Big|  \\
&\leq \eta \int_{0}^t \int_\mathbb{R} e^{2\mu s}\omega(\xi)
 |f'(\phi(\xi-cr))|u^2(s,\xi)\,d\xi\,ds \\
&\quad +\frac{1}{\eta}\int_{0}^t \int_\mathbb{R} e^{2\mu s}\omega(\xi)
 |f'(\phi(\xi-cr))| u^2(s-r,\xi-cr)\,d\xi\,ds \\
&=\eta \int_{0}^t \int_\mathbb{R} e^{2\mu s}\omega(\xi)|f'(\phi(\xi-cr))|
  u^2(s,\xi)\,d\xi\,ds \\
&\quad +\frac{1}{\eta}e^{2\mu r}\int_{-r}^{t-r}
 \int_\mathbb{R} e^{2\mu s}\omega(\xi+cr)|f'(\phi(\xi))|u^2(s,\xi)\,d\xi\,ds \\
&\leq \eta \int_{0}^t \int_\mathbb{R} e^{2\mu s}\omega(\xi)|f'(\phi(\xi-cr))|
 u^2(s,\xi)\,d\xi\,ds \\
&\quad +\frac{1}{\eta}e^{2\mu r}\int_{0}^{t}
 \int_\mathbb{R} e^{2\mu s}\omega(\xi+cr)|f'(\phi(\xi))|u^2(s,\xi)\,d\xi\,ds \\
&\quad +\frac{1}{\eta}e^{2\mu r}\int_{-r}^{0}
 \int_\mathbb{R} e^{2\mu s}\omega(\xi+cr)|f'(\phi(\xi))|u_0^2(s,\xi)\,d\xi\,ds,
\end{aligned}
\end{equation}
Substituting \eqref{4.7} in \eqref{4.6}, we have
\begin{equation}\label{4.8}
\begin{aligned}
&e^{2\mu t}\|u(t)\|_{{L_\omega^2}}^2+\int_{0}^t
 \int_\mathbb{R}e^{2\mu s} B_{\eta,\mu,\omega}(\xi)\omega(\xi)u^2(s,\xi)\,d\xi\,ds  \\
&\leq \|u_0(0)\|_{{L_\omega^2}}^2+\frac{e^{2\mu r}}{\eta}
 \int_{-r}^0 \int_\mathbb{R}e^{2\mu s}\omega(\xi+cr)|f'(\phi(\xi)|u_0^2(s,\xi)\,d\xi\,ds  \\
&\quad +2\int_{0}^t\int_\mathbb{R} e^{2\mu s} \omega(\xi) u(s,\xi)
 [G(u)(s,\xi)-E(u)(s,\xi)]\,d\xi\,ds,
\end{aligned}
\end{equation}
where $B_{\eta,\mu,\omega}(\xi)$ is given by \eqref{4.2}.

By  standard Sobolev's embedding inequality
$H^1(\mathbb{R})\hookrightarrow C^0(\mathbb{R})$ and the embedding inequality
 $H_\omega^1(\mathbb{R})\hookrightarrow H^1(\mathbb{R})$
(since $\omega(\xi)\geq1$, for all $\xi\in \mathbb{R}$), we have, for
all $\xi\in \mathbb{R}$  and $-r\leq t \leq T$,
\begin{equation}\label{4.9}
|u(t,\xi)|\leq\sup_{\xi\in \mathbb{R}}|u(t,\xi)|
\leq\sigma_0\|u(t,\cdot)\|_{H^1}\leq\sigma_0\|u(t,\cdot)\|_{H_\omega^1}\leq\sigma_0M(t),
\end{equation}
where $\sigma_0>0$ is the embedding constant. Since
\begin{align*}
|G(u)(t,\xi)|
&=|f(u(t-r,\xi-cr)+\phi(\xi-cr))-f(\phi(\xi-cr))\\
&\quad -f'(\phi(\xi-cr))u(t-r,\xi-cr)|\leq C|u(t-r,\xi-cr)|^2,
\end{align*}
we obtain
\begin{equation}\label{4.10}
\begin{aligned}
&2\Big|\int_{0}^t\int_\mathbb{R} e^{2\mu s} \omega(\xi)
  u(s,\xi)G(u)(s,\xi)\,d\xi\,ds\Big| \\
&\leq C\sigma_{0}M(t)\int_{0}^t\int_\mathbb{R} e^{2\mu s} \omega(\xi)u^2(s-r,\xi
 -cr)\,d\xi\,ds \\
&=C\sigma_{0}M(t)\int_{-r}^{t-r}\int_\mathbb{R} e^{2\mu (s+r)}
 \omega(\xi+cr) u^2(s,\xi)\,d\xi\,ds \\
&\leq CM(t)e^{2\mu r}\Big\{\int_{0}^t\int_\mathbb{R} e^{2\mu s}
 \omega(\xi)u^2(s,\xi)\,d\xi\,ds+\int_{-r}^0\int_\mathbb{R} e^{2\mu s}
 \omega(\xi+cr)u^2_0(s,\xi)\,d\xi\,ds\Big\} \\
&\leq CM(t)\Big\{\int_{0}^t\int_\mathbb{R} e^{2\mu s} \omega(\xi)u^2
 (s,\xi)\,d\xi\,ds+\int_{-r}^0\int_\mathbb{R} e^{2\mu s}
 \omega(\xi+cr)u^2_0(s,\xi)\,d\xi\,ds\Big\}.
\end{aligned}
\end{equation}
On the other hand,
$$
|E(u)(t,\xi)|=|d(u(t,\xi)+\phi(\xi))-d(\phi(\xi))-d'(\phi(\xi))u(t,\xi)|
\leq C|u(t,\xi)|^2,
$$
we can also obtain
\begin{equation}\label{4.11}
2\Big|\int_{0}^t\int_\mathbb{R} e^{2\mu s} \omega(\xi) u(s,\xi)E(u)(s,\xi)
 \,d\xi\,ds\Big|
\leq CM(t)\int_{0}^t\int_\mathbb{R} e^{2\mu s} \omega(\xi)u^2(s,\xi)\,d\xi\,ds.
\end{equation}
From \eqref{4.10}, \eqref{4.11} and \eqref{4.8}, we have
\begin{equation}\label{4.12}
\begin{aligned}
&e^{2\mu t}\|u(t)\|_{L_{\omega}^2}^2+\int_{0}^t
 \int_\mathbb{R}e^{2\mu s} [B_{\eta,\mu,\omega}(\xi)-CM(t)]
 \omega(\xi)u^2(s,\xi)\,d\xi\,ds  \\
&\leq \|u_0(0)\|_{L_{\omega}^2}^2+\int_{-r}^0
 \int_\mathbb{R}e^{2\mu s}\omega(\xi+cr)[CM(t)
 +\frac{e^{2\mu r}}{\eta} |f'(\phi(\xi)|]u_0^2(s,\xi)\,d\xi\,ds \\
&\leq C\Big(\|u_0(0)\|_{L_{\omega}^2}^2
 +\int_{-r}^0 \|u_0(s)\|_{L_{\omega}^2}^2 ds\Big),
\end{aligned}
\end{equation}
which immediately implies \eqref{4.1}. This completes the proof.
\end{proof}

Next we prove a key inequality.

\begin{lemma}\label{lem4.2}
 Letting $\eta=e^{-\lambda cr}$, there exists a unique number $c_\ast>0$,
such for every $c>c_\ast$, there exists a constant $C_1>0$ such that
\begin{equation}\label{4.13}
A_{\eta,\omega}(\xi)\geq C_1>0  \quad \text{for }  \xi \in \mathbb{\mathbb{R}}.
\end{equation}
\end{lemma}

\begin{proof}
We distinguish three cases:
\smallskip

\noindent\textbf{Case 1:}
 For $ \xi<\xi_\ast-cr,$ $ \omega(\xi)=e^{-2\lambda(\xi-\xi_\ast)}$ and
$\omega(\xi+cr)=e^{-2\lambda(\xi-\xi_\ast+cr)}$
$$
\frac {\omega'(\xi)}{\omega(\xi)}=-2\lambda,\quad
\frac {\omega(\xi+cr)}{\omega(\xi)}=e^{-2\lambda cr}.
$$
By  (H4) and \eqref{1.7}, we can have

\begin{align}
A_{\eta,\omega}(\xi)
&:=-c\frac{\omega'(\xi)} {\omega(\xi)}+2d'(\phi(\xi))-\frac{D}{2}
 \Big(\frac{ \omega'(\xi)}{\omega(\xi)}\Big)^2
 -\eta |f'(\phi(\xi-cr))| \nonumber \\
&\quad -\frac{1}{\eta}|f'(\phi(\xi))|
 \frac{\omega(\xi+cr)} {\omega(\xi)}  \label{4.14} \\
&=2c\lambda-2D\lambda^2+2d'(\phi(\xi))-e^{-\lambda cr}
 |f'(\phi(\xi-cr))|-e^{-\lambda cr}|f'(\phi(\xi))| \nonumber \\
&\geq2\Big(c\lambda-D\lambda^2+d'(0)-e^{-\lambda cr}f'(0)\Big) 
=:C_{11}>0. \nonumber
\end{align}
\smallskip

\noindent\textbf{Case 2:} For $\xi_\ast-cr\leq\xi\leq\xi_\ast$, then
$ \omega(\xi)=e^{-2\lambda(\xi-\xi_\ast)}$ and $\omega(\xi+cr)=1$,
 and by Lemma \ref{lem2.2}, we have
\begin{equation}\label{4.15}
\begin{aligned}
&A_{\eta,\omega}(\xi)\\
&:=-c\frac{\omega'(\xi)} {\omega(\xi)}+2d'(\phi(\xi))
 -\frac{D}{2}\Big(\frac{ \omega'(\xi)}{\omega(\xi)}\Big)^2
 -\eta |f'(\phi(\xi-cr))| \\
&\quad -\frac{1}{\eta}|f'(\phi(\xi))|  \frac{\omega(\xi+cr)} {\omega(\xi)} \\
&=2c\lambda-2D\lambda^2+2d'(\phi(\xi))-\eta|f'(\phi(\xi-cr))| \\
&\quad  -\frac{1}{\eta}|f'(\phi(\xi))|e^{2\lambda(\xi-\xi_\ast)} \\
&\geq2(c\lambda-D\lambda^2+d'(0)-e^{-\lambda cr}f'(0))
 +2e^{-\lambda cr}f'(0)-\eta|f'(\phi(\xi-cr))| \\
&\quad -\frac{1}{\eta}|f'(\phi(\xi))| \\
&\geq2\triangle(c,\lambda)+2e^{-\lambda cr}f'(0)-(\eta+\frac{1}{\eta})
 \max\{|f'(\phi(\xi-cr))|,|f'(\phi(\xi))|\} \\
&\geq2\triangle(c,\lambda)+2e^{-\lambda cr}f'(0)-2\max\{|f'(\phi(\xi-cr))|,
 |f'(\phi(\xi))|\}\cosh(\lambda cr) \\
&\geq2\Big(\triangle(c,\lambda)+e^{-\lambda cr}f'(0)
 -\max\{|f'(\phi(\xi_0-cr))|,|f'(\phi(\xi_0))|\}\cosh(\lambda cr)\Big) \\
&=:C_{12}>0.
\end{aligned}
\end{equation}
\smallskip

\noindent\textbf{Case 3:}
 For $ \xi\geq\xi_\ast, \,\omega(\xi)=\omega(\xi+cr)=1$, and by Lemma \ref{lem2.2},
we obtain
\begin{equation}\label{4.16}
\begin{aligned}
A_{\eta,\omega}(\xi):
&=2d'(\phi(\xi))-\eta|f'(\phi(\xi-cr))|-\frac{1}{\eta}|f'(\phi(\xi))| \\
&\geq2\Big(d'(0)-(\eta+\frac{1}{\eta})\max\{|f'(\phi(\xi-cr))|,
 |f'(\phi(\xi))|\}\Big) \\
&\geq2\Big(d'(0)-\max\{|f'(\phi(\xi_0-cr))|,|f'(\phi(\xi_0))|\}
 \cosh(\lambda cr)\Big) \\
&=:C_{13}>0.
\end{aligned}
\end{equation}
Combining \eqref{4.14}--\eqref{4.16}, we obtain
 $A_{\eta,\omega}(\xi)\geq C_1,$
where $C_1=\min_{i=1,2,3}\{C_{1i}\}>0$. This completes the proof.
\end{proof}

\begin{lemma}\label{lem4.3}
Let $u(t,\xi)\in X(-r,T)$. Then there exists a constant $\mu^*>0$ such that
for $0<\mu<\mu^*$, it holds
\begin{equation}\label{4.17}
\begin{aligned}
&\|u(t)\|_{{L_\omega^2}}^2+\int_{0}^t
 \int_\mathbb{R}e^{-2\mu (t-s)}\|u(s)\|_{{L_\omega^2}}^2 ds\\
&\leq Ce^{-2\mu t}\Big(\|u_0(0)\|_{L^2_{\omega}}^2+\int_{-r}^0
 \|u_0(s)\|_{{L_\omega^2}}^2  ds\Big ),
\end{aligned}
\end{equation}
provided $M(t)\ll 1$.
\end{lemma}

\begin{proof}
 We distinguish three cases:
\smallskip

\noindent\textbf{Case 1:} For $ \xi<\xi_\ast-cr$,
$ \omega(\xi)=e^{-2\lambda(\xi-\xi_\ast)}$,
$\omega(\xi+cr)=e^{-2\lambda(\xi-\xi_\ast+cr)}$,
and according to Lemma \ref{lem4.2},
\begin{equation}\label{4.18}
\begin{aligned}
B_{\eta,\mu,\omega}(\xi)£º&=A_{\eta,\omega}(\xi)-2\mu-\frac{1}{\eta}(e^{2\mu r}-1)|f'(\phi(\xi))|\frac{\omega(\xi+cr)} {\omega(\xi)} \\
&\geq C_1-2\mu-f'(0)e^{-\lambda cr}(e^{2\mu r}-1) \\
&=:C_{21}>0 \,\,\text{for }\,\,0<\mu<\mu_1,
\end{aligned}
\end{equation}
where $\mu_1$ is the unique root of the equation
\[
C_1-2\mu-f'(0)e^{-\lambda cr}(e^{2\mu r}-1)=0.
\]
\smallskip

\noindent\textbf{Case 2:}
 For $\xi_\ast-cr\leq\xi\leq\xi_\ast$, then
 $ \omega(\xi)=e^{-2\lambda(\xi-\xi_\ast)}$ and $\omega(\xi+cr)=1$,
\begin{equation}\label{4.19}
\begin{aligned}
B_{\eta,\mu,\omega}(\xi)
&=A_{\eta,\omega}(\xi)-2\mu-\frac{1}{\eta}(e^{2\mu r}-1)|f'(\phi(\xi))|
 \frac{\omega(\xi+cr)} {\omega(\xi)} \\
&\geq C_1-2\mu-f'(0)e^{\lambda cr}(e^{2\mu r}-1)e^{2\lambda(\xi-\xi_*)} \\
&\geq C_1-2\mu-f'(0)e^{\lambda cr}(e^{2\mu r}-1) \\
&=:C_{22}>0 \quad \text{for } 0<\mu<\mu_2
\end{aligned}
\end{equation}
where $\mu_2$ is the unique root of the equation
\[
C_1-2\mu-f'(0)e^{\lambda cr}(e^{2\mu r}-1)=0.
\]
\smallskip

\noindent\textbf{Case 3:}
 For $ \xi\geq\xi_\ast, \omega(\xi)=\omega(\xi+cr)=1$,
\begin{equation}\label{4.20}
\begin{aligned}
B_{\eta,\mu,\omega}(\xi)
&=A_{\eta,\omega}(\xi)-2\mu-\frac{1}{\eta}(e^{2\mu r}-1)|f'(\phi(\xi))|
 \frac{\omega(\xi+cr)} {\omega(\xi)} \\
&\geq C_1-2\mu-f'(0)e^{\lambda cr}(e^{2\mu r}-1) \\
&=:C_{22}>0 \quad\text{for } 0<\mu<\mu_2.
\end{aligned}
\end{equation}
Combining \eqref{4.18}, \eqref{4.19} and \eqref{4.20}, we  obtain
$B_{\eta,\mu,\omega}(\xi)\geq C_2,$ for $0<\mu<\mu_*$, where
$C_2:=\min\{C_{21},C_{22}\}$ and $\mu_*=\min\{\mu_1,\mu_2\}$.
It follows from \eqref{4.1} that
\begin{align*}
&\|u(t)\|_{{L_\omega^2}}^2+\int_{0}^t e^{-2\mu (t-s)}
 \int_\mathbb{R}  [C_2-CM(t)]\omega(\xi)u^2(s,\xi)\,d\xi\,ds\\
&\leq Ce^{-2\mu t}\Big(\|u_0(0)\|_{L^2_{\omega}}^2
 +\int_{-r}^0 \|u_0(s)\|_{{L_\omega^2}}^2 \,d\xi\,ds\Big ),
\end{align*}
which implies \eqref{4.17} by letting $M(t)\ll1$. This completes the proof.
\end{proof}

Next, we shall establish the energy estimate for $u_\xi$,
 which is similar to \eqref{4.17}.

\begin{lemma}\label{lem4.4}
Let $u(t,\xi)\in X(-r,T)$. Then it holds
\begin{equation}\label{4.210}
\begin{aligned}
&\|u_\xi(t)\|_{{L_\omega^2}}^2+\int_{0}^t e^{-2\mu (t-s)}
 \|u_\xi(t)\|_{{L_\omega^2}}^2 ds \\
&\leq Ce^{-2\mu t}\Big(\|u_0(0)\|_{H^1_{\omega}}^2
+\int_{-r}^0 \|u_0(s)\|_{{H_\omega^1}}^2 ds \Big).
\end{aligned}
\end{equation}
provided $M(t)\ll 1$.
\end{lemma}

\begin{proof}
 Differentiating \eqref{3.1} with respect to $\xi$ and multiplying the
obtained equation by $e^{2\mu t}\omega(\xi)u_\xi(t,\xi)$, we have
\begin{equation}\label{4.21}
\begin{aligned}
&\Big\{\frac{1}{2}e^{2\mu t}\omega u_\xi^2\Big\}_t
 +e^{2\mu t}\Big\{\frac{1}{2}c\omega u_\xi^2-D\omega u_\xi u_{\xi\xi}\Big\}_\xi
 +De^{2\mu t}\omega u_{\xi\xi}^2+De^{2\mu t}\omega'u_{\xi}u_{\xi\xi} \\
&+\Big\{-\frac{c}{2}\frac{\omega'}{\omega}+d'(\phi(\xi))-\mu\Big\}
 e^{2\mu t}\omega u_\xi^2-e^{2\mu t}\omega(\xi) u_\xi(t,\xi)
 f'(\phi(\xi-cr)) \\
&\quad\times u_\xi(t-r,\xi-cr) \\
&=e^{2\mu t}\omega(\xi) u_\xi(t,\xi)[G_2(u)+G_1(u)]-e^{2\mu t}\omega(\xi)
 u_\xi(t,\xi)[E_2(u)+E_1(u)],
\end{aligned}
\end{equation}
where
\begin{gather*}
G_1(u)(t,\xi)=[f'(u(t-r,\xi-cr)+\phi(\xi-cr))-f'(\phi(\xi-cr))]\phi'(\xi-cr),\\
G_2(u)(t,\xi)=[f'(u(t-r,\xi-cr)+\phi(\xi-cr))-f'(\phi(\xi-cr))]u_\xi(t-r,\xi-cr),\\
E_1(u)(t,\xi)=[d'(u(t,\xi)+\phi(\xi))-d'(\phi(\xi))]\phi'(\xi),\\
E_2(u)(t,\xi)=[d'(u(t,\xi)+\phi(\xi))-d'(\phi(\xi))]u_\xi(t,\xi).\\
\end{gather*}
Using the Cauchy-Schwarz inequality
$$
|De^{2\mu t}\omega'u_{\xi}u_{\xi\xi}|
 \leq De^{2\mu t}\omega u_{\xi\xi}^2+\frac {D}{4}e^{2\mu t}
 \Big(\frac{\omega'}{\omega}\Big)^2\omega u_\xi^2,
$$
it follows from \eqref{4.21} that
\begin{equation}\label{4.22}
\begin{aligned}
&\Big\{\frac{1}{2}e^{2\mu t}\omega u_\xi^2\Big\}_t
 +e^{2\mu t}\Big\{\frac{1}{2}c\omega u_\xi^2-D\omega u_\xi u_{\xi\xi}\Big\}_\xi \\
&+\Big\{-\frac{c}{2}\frac{\omega'}{\omega}+d'(\phi(\xi))-\mu
 -\frac {D}{4}\Big(\frac{\omega'}{\omega}\Big)^2\Big\}e^{2\mu t}\omega u_\xi^2 \\
&-e^{2\mu t}\omega(\xi) u_\xi(t,\xi) f'(\phi(\xi-cr))u_\xi(t-r,\xi-cr) \\
&\leq e^{2\mu t}\omega(\xi) u_\xi(t,\xi)[G_2(u)+G_1(u)]
 -e^{2\mu t}\omega(\xi) u_\xi(t,\xi)[E_2(u)+E_1(u)].
\end{aligned}
\end{equation}
Integrating the above inequality over $\mathbb{R}\times[0,t]$
with respect to $\xi$ and $ t $, we have
\begin{align}
&e^{2\mu t}\|u_\xi(t)\|_{{L_\omega^2}}^2
 +\int_{0}^t \int_\mathbb{R} e^{2\mu s}
 \Big\{-c\frac{\omega'(\xi)}{\omega(\xi)}+2d'(\phi(\xi))
 -2\mu-\frac {D}{2}\Big(\frac{\omega'(\xi)}{\omega(\xi)}\Big)^2\Big\} \nonumber \\
&\quad\times \omega(\xi)u_\xi^2(s,\xi)\,d\xi\,ds  \nonumber \\
&\quad -2\int_{0}^t \int_\mathbb{R} e^{2\mu s}\omega (\xi)f'(\phi(\xi-cr))
 u_\xi(s,\xi) u_\xi(s-r,\xi-cr)\,d\xi\,ds  \label{4.23} \\
&\leq \|u_0(0)\|_{H_{\omega}^1}^2
+2\int_{0}^t\int_\mathbb{R} e^{2\mu s} \omega(\xi) u_\xi(s,\xi)
 [G_2(u)(s,\xi)+G_1(u)(s,\xi)]\,d\xi\,ds \nonumber\\
&\quad -2\int_{0}^t\int_\mathbb{R} e^{2\mu s} \omega(\xi) u_\xi(s,\xi)
 [E_2(u)(s,\xi)+E_1(u)(s,\xi)]\,d\xi\,ds. \nonumber
\end{align}
By the Cauchy-Schwarz inequality, we have
\begin{equation}\label{4.24}
\begin{aligned}
&2\Big|\int_{0}^t \int_\mathbb{R} e^{2\mu s}\omega (\xi)f'(\phi(\xi-cr))u_\xi(s,\xi)
 u_\xi(s-r,\xi-cr)\,d\xi\,ds \Big|  \\
&\leq \eta \int_{0}^t \int_\mathbb{R} e^{2\mu s}\omega(\xi)|
 f'(\phi(\xi-cr))|u_\xi^2(s,\xi)\,d\xi\,ds \\
&\quad +\frac{1}{\eta}e^{2\mu r}\int_{0}^t \int_\mathbb{R} e^{2\mu s}
 \omega(\xi+cr)|f'(\phi(\xi))|u_{\xi}^2(s,\xi)\,d\xi\,ds \\
&\quad +\frac{1}{\eta}e^{2\mu r}\int_{-r}^0 \int_\mathbb{R} e^{2\mu s}
 \omega(\xi+cr)|f'(\phi(\xi))|u_{0\xi}^2(s,\xi)\,d\xi\,ds.
\end{aligned}
\end{equation}
It follows from \eqref{4.23} and \eqref{4.24} that
\begin{equation}\label{4.25}
\begin{aligned}
&e^{2\mu t}\|u_\xi(t)\|_{{L_\omega^2}}^2
+\int_{0}^t \int_\mathbb{R} e^{2\mu s}B_{\eta,\mu,\omega}(\xi)
 \omega(\xi)u_\xi^2(s,\xi)\,d\xi\,ds  \\
&\leq \|u_0(0)\|_{H_{\omega}^1}^2+\frac{e^{2\mu r}}{\eta}\int_{-r}^0
 \int_\mathbb{R}e^{2\mu s}\omega(\xi+cr) |f'(\phi(\xi))|u_{0\xi}^2(s,\xi)\,d\xi\,ds \\
&\quad +2\int_{0}^t\int_\mathbb{R} e^{2\mu s} \omega(\xi) u_\xi(s,\xi)
 [G_2(u)(s,\xi)+G_1(u)(s,\xi)]\,d\xi\,ds \\
&\quad -2\int_{0}^t\int_\mathbb{R} e^{2\mu s} \omega(\xi) u_\xi(s,\xi)
 [E_2(u)(s,\xi)+E_1(u)(s,\xi)]\,d\xi\,ds.
\end{aligned}
\end{equation}
Again, by the Taylor expansion,
\begin{align*}
|G_2(u)(t,\xi)|
&=\Big|[f'(u(t-r,\xi-cr)+\phi(\xi-cr))-f'(\phi(\xi-cr))]\Big|u_\xi(t-r,\xi-cr)\\
&\leq C|u(t-r,\xi-cr)|u_\xi(t-r,\xi-cr),
\end{align*}
we can estimate the nonlinear term as
\begin{align}
&2\Big|\int_{0}^t\int_\mathbb{R} e^{2\mu s} \omega(\xi)
 u_\xi(s,\xi) G_2(u)(s,\xi)\,d\xi\,ds\Big| \nonumber \\
&\leq C\int_{0}^t\int_\mathbb{R} e^{2\mu s} \omega(\xi) u_\xi(s,\xi)
 |u(s-r,\xi-cr)|u_\xi(s-r,\xi-cr)\,d\xi\,ds \nonumber \\
&\leq CM(t)\int_{0}^t\int_\mathbb{R} e^{2\mu s} \omega(\xi) u_\xi^2(s,\xi)
 \,d\xi\,ds  \label{4.26} \\
&\quad +CM(t)\Big(\int_{-r}^0\int_\mathbb{R}
+\int_{0}^t\int_\mathbb{R}\Big) e^{2\mu (s+r)} 
 \omega(\xi+cr) u_\xi^2(s,\xi)\,d\xi\,ds \nonumber \\
&\leq  CM(t)\int_{0}^t\int_\mathbb{R} e^{2\mu s} \omega(\xi)
 u_\xi^2(s,\xi)\,d\xi\,ds \nonumber \\
&\quad +CM(t)\int_{-r}^0\int_\mathbb{R} e^{2\mu (s+r)} \omega(\xi+cr)
 u_{0\xi}^2(s,\xi)\,d\xi\,ds, \nonumber
\end{align}
and
\begin{equation}\label{4.27}
\begin{aligned}
&2\Big|\int_{0}^t\int_\mathbb{R} e^{2\mu s} \omega(\xi) u_\xi(s,\xi)
  G_1(u)(s,\xi)\,d\xi\,ds\Big| \\
&\leq C\Big|\int_{0}^t\int_\mathbb{R} e^{2\mu s} \omega(\xi) u_\xi(s,\xi)
 u(s-r,\xi-cr)\phi'(\xi-cr)\,d\xi\,ds\Big| \\
&\leq C\int_{0}^t\int_\mathbb{R} e^{2\mu s} \omega(\xi) u_\xi(s,\xi)
 |u(s-r,\xi-cr)|\,d\xi\,ds \\
&\leq C\int_{0}^t\int_\mathbb{R} e^{2\mu s} \omega(\xi) u_\xi^2(s,\xi)\,d\xi\,ds \\
&\quad +C\Big(\int_{-r}^0\int_\mathbb{R}+\int_{0}^t\int_\mathbb{R}\Big) e^{2\mu (s+r)}
  \omega(\xi+cr) u^2(s,\xi)\,d\xi\,ds.
\end{aligned}
\end{equation}
Similarly, we obtain
\[ %\label{4.28}
2\Big|\int_{0}^t\int_\mathbb{R} e^{2\mu s} \omega(\xi)
u_\xi(s,\xi) E_2(u)(s,\xi))\,d\xi\,ds\Big|
\leq CM(t)\int_{0}^t\int_\mathbb{R} e^{2\mu s} \omega(\xi) u_\xi^2(s,\xi)\,d\xi\,ds
\]
and
\begin{equation}\label{4.29}
\begin{aligned}
&2\Big|\int_{0}^t\int_\mathbb{R} e^{2\mu s} \omega(\xi) u_\xi(s,\xi)
 E_1(u)(s,\xi))\,d\xi\,ds\Big| \\
&\leq C\Big|\int_{0}^t\int_\mathbb{R} e^{2\mu s} \omega(\xi) u_\xi(s,\xi)
u(s,\xi))\phi'(\xi-cr)\,d\xi\,ds\Big| \\
&\leq C\int_{0}^t\int_\mathbb{R} e^{2\mu s} \omega(\xi) u_\xi^2(s,\xi)\,d\xi\,ds \\
&\quad +C\int_{0}^t\int_\mathbb{R} e^{2\mu s} \omega(\xi) u^2(s,\xi)\,d\xi\,ds.
\end{aligned}
\end{equation}
It then follows from \eqref{4.26}-\eqref{4.29} and Lemma \ref{lem4.3} that
\[
\|u_\xi(t)\|_{{L_\omega^2}}^2+\int_{0}^t e^{-2\mu (t-s)}
 \|u_\xi(t)\|_{{L_\omega^2}}^2 ds
\leq Ce^{-2\mu t}\Big(\|u_0(0)\|_{H^1_{\omega}}^2
 +\int_{-r}^0 \|u_0(s)\|_{{H_\omega^1}}^2  ds\Big ).
\]
Combining \eqref{4.17} and \eqref{4.210}, for some constant $C$, which is
independent of $T$ and $u(t,\xi)$, we have
$$
\|u(t)\|_{H^1_{\omega}}^2 \leq Ce^{-2\mu t}
\Big(\|u_0(0)\|_{H^1_{\omega}}^2+\int_{-r}^0 \|u_0(s)\|_{{H_\omega^1}}^2 ds\Big),
\quad \text{for  all } 0\leq t\leq T.
$$
This completes the proof.
\end{proof}


\begin{proof}[Proof of Theorem \ref{thm2.1}]
It is based on the existence of a local solution and the estimate obtained above.
The process is similar to the one in \cite{Mei-So,Mei-So-Li}, using the
continuity extension method, so we omit it.
\end{proof}

\subsection*{Acknowledgments}
The second author was supported by the National Natural
Science Foundation of China, by Shanghai Leading Academic Discipline
Project (No. XTKX2012), by Innovation Program of Shanghai Municipal Education
Commission (No. 14YZ096), by the Hujiang Foundation of China (B14005).
The third author was supported by National Natural Science Foundation of
China (No. 11301542).

\begin{thebibliography}{10}

\bibitem{Aguerrea-Gomez-Trofimchuk}
 M. Aguerrea, C. Gomez, S. Trofimchuk;
On uniqueness of semi-wavefronts, \emph{Math. Ann.,} \textbf{354} (2012), 73-109.

\bibitem{Aguerrea} M. Aguerrea;
 On the uniqueness of semi-wavefronts for non-local delayed reaction-diffusion
equations, \emph{J. Math. Anal. Appl.,} \textbf{62} (2011), 377-397.

\bibitem{Chern-Mei} I. Chern, M. Mei, X. Yang, Q. Zhang;
 Stability of no-monotone critical traveling waves for reaction-diffusion
equation with time-delay, \emph{J. Diff. Eqns.}, 259 (2015), 1503-1541.

\bibitem{Hung} D. Duehring, W. Huang;
 Periodic traveling waves for diffusion equation with time delayed and non-local
responding reaction. \emph{J. Diff. Dyn. Eqns,} \textbf{19} (2007), 457-477.

\bibitem{Fang-Zhao}J. Fang, X. Zhao;
Existence and uniqueness of traveling waves for non-monotone integral equations
 with applications, \emph{J. Diff. Eqns,} \textbf{248} (2010), 2199-2226.

\bibitem{Faria-Huang-Wu} T. Faria, W. Huang, J. Wu;
Traveling waves for delayed reaction-diffusion equations with global response,
 \emph{Proc.  Roy. Soc. London,} \textbf{462} (2006), 229-261.

\bibitem{Faria-Trofimchuk} T. Faria, S. Trofimchuk;
 Nonmontone traveling waves in a single species reaction-diffusion equation
with delay, \emph{J. Diff. Eqns,} \textbf{288} (2006), 357-376.

\bibitem{Faria-Trofimchuk1} T. Faria, S. Trofimchuk;
 Positive traveling fronts for reaction-diffusion systems with distributed delay,
\emph{Nonlinearity,} \textbf{23} (2010),2457-2481.

\bibitem{Gomez-Trofimchuk} A. Gomez, S. Trofimchuk;
Monotone traveling wavefronts of the KPP-Fisher delayed equation,
\emph{J. Diff. Eqns,} \textbf{250} (2011), 1767-1787.

\bibitem{Gomez-Trofimchuk1} A. Gomez, S. Trofimchuk;
 Global continucation of monotone wavefronts of the KPP-Fisher delayed
equation, \emph{J. London Math. Soc.,} \textbf{89} (2014), 47-68.

 \bibitem{Gourley-So-Wu} S. Gourley, J. So, J. Wu;
Nonlocality of reaction-diffusion equations induced
by delay: Biological modeling and nonlinear dynamics,
in Contemporary Mathematics, Thematic Surveys, D. V. Anosov and A. Skubachevskii,
eds., Kluwer Plenum, 2003, pp. 84¨C
120 (in Russian); \emph{J. Math. Sci.}, \textbf{124} (2004),
5119-5153 (in English).

 \bibitem{Gourley-Wu} S. A. Gourley, J. Wu;
Delayed nonlocal diffusive system in biological invasion and disease
spread, \emph{Fields Inst. Commun.}, \textbf{48} (2006), 137-200.

\bibitem{Lin-Mei} C. K. Lin,  M. Mei;
 On traveling wavefronts of the Nicholson's blowflies equation with diffusion,
\emph{Proc.  Roy. Soc. Edinb.,} \textbf{140} (2010), 135-152.

\bibitem{Chi-Mei} C. K. Lin, C. T. Lin, Y. Lin, M. Mei;
Exponential stability of nonmonotone traveling waves for Nicholson's
 Blowfilws equation,\emph{SIAM. J. Math. Anal.,} \textbf{46} (2014), 1053-1084.

\bibitem{Lv-Wang} G. Lv, M. Wang;
Nonlinear stability of traveling wave fronts for delayed reaction diffusion,
\emph{Proc.  Roy. Soc. Edinb.,} \textbf{140} (2010), 1609-1630.

\bibitem{Lv-Wang1} G. Lv, M. Wang;
Nonlinear stability of travelling wave fronts for delayed reaction
diffusion equations, \emph{Nonlinearity,} \textbf{23} (2010), 845-873.

\bibitem{Ma} S. Ma;
Traveling waves for non-local delayed diffusion equations via auxiliary equations,
 \emph{J. Diff. Eqns,} \textbf{237} (2007), 259-277.

\bibitem{Matsumura-Nishida} A. Matsumura, T. Nishida;
The initial value problem for the equations of motion of viscous and
heat-conductive gases,\emph{J. Math. Kyoto. Univ.} \textbf{20} (1980), pp.67-104.

\bibitem{Mei} M. Mei;
 Global smooth solutions of the cauchy problem for higher-dimensional
generalized pulse transmission equations, \emph{Acta. Math.  Appl. Sin.,}
\textbf{14} (1991), 450-461.

\bibitem{Mei-So} M. Mei, J.  So;
 Stability of strong traveling waves for a nonlocal time-delayed
reaction-diffusion equation, \emph{Proc. Roy. Soc. Edinb.} \emph{138} (2008),
551-568.

\bibitem{Mei-So-Li} M. Mei, J. So, M. Li, S. Shen;
Asymptotic stability of traveling waves for the Nicholson's blowflies equation
with diffusion, \emph{Proc. Roy. Soc. Edinb.,} \textbf{134} (2004), 579-594.

\bibitem{Mei1} M. Mei, C. K. Lin, C. T. Lin, J. So;
Traveling wavefronts for time-delayed reaction-diffusion equation:
(I) local nonlinearity, \emph{J. Diff. Eqns,} \textbf{247} (2009), 495-510.

\bibitem{Mei2} M. Mei, C. K. Lin, C. T. Lin, J. So;
 Traveling wavefronts for time-delayed reaction-diffusion equation:
(II) nonlocal nonlinearity, \emph{J. Diff. Eqns,} \textbf{247} (2009), 511-529.

\bibitem{Mei-Ou-Zhao} M. Mei, C. Ou, X. Zhao;
 Global stability pf monostable traveling waves for nonlocal time-delayed
reaction-diffusion equations, \emph{SIAM J.  Math. Anal.,} \textbf{42} (2010),
2762-2790.

\bibitem {Mei-Wang} M. Mei, Y. Wang;
Remark on stability of traveling waves for nonlocal Fisher-KPP equations,
\emph{Int J. Num. Anal. Model. B,} \textbf{2} (2011), 379-401.


\bibitem{Schaaf} K. Schaaf;
 Asymptotic behavior and traveling wave solution for parabolic functional
differential equations, \emph{Trans. Amer. Math. Soc.,} \textbf{302} (1987), 587-615.

\bibitem{So-Yang} J. So, Y. Yang;
Dirichlet problem for the diffusive Nicholson¡¯s blowflies equation,
\emph{J. Diff. Eqns.},  \textbf{150} (1998), 317-348.

\bibitem{So-Zou} J. So, X. Zou;
 Traveling waves for the diffusive Nicholson's blowflies equation,
\emph{Appl. Math. Comput.,} \textbf{122} (2001), 32-42.

\bibitem{Trofimchuk-Pinto} E. Trofimchuk, M. Pinto, S. Trofimchuk;
 Pushed traveling fronts in monostable equations with monotone delayed reaction,
\emph{Discrete Contin. Dyn. Syst.,} \textbf{33} (2013), 2169-2187.

\bibitem{Wu-Li-Liu} S. Wu, W. Li, S. Liu;
 Oscillatory waves in reaction-diffusion equations with nonlocal delay and
crossing-monostability. \emph{Nonlinear Anal.  RWA}, \textbf{10} (2009), 3141-3151.

\bibitem{Wu-Li-Liu2} S. Wu, W. Li,  S. Liu;
Asymptotic stability of traveling wave fronts in nonlocal reaction-diffusion
equations with delay, \emph{J.  Math. Anal. Appl.,} \textbf{360} (2009), 439-458.


\bibitem{Wang-Li} Z. Wang, W. Li;
 Dynamics of a non-local delayed reaction-diffusion equation without
quasi-monotonicity, \emph{Proc. Roy. Soc. Edinb.,}  \textbf{140} (2010), 1081-1109.

\bibitem{Wu-Liu} S. Wu, S. Liu;
Existence and uniqueness of traveling waves for non-monotone integral
equations with application, \emph{J. Math. Anal. Appl.}, \textbf{365} (2010), 729-741

\bibitem{Wu-Zhao-Liu} S. Wu, W. Zhao, S. Liu;
 Asymptoic stability of traveling waves for delayed reaction-diffusion
equations with crossing-monostability, \emph{Z. Angew. Math. Phys.},
 \textbf{62} (2011), 377-397.

\end{thebibliography}

\end{document}
