A multi-dimensional bistable nonlinear diffusion equation in a periodic medium (Q330879): Difference between revisions

The author studies the existence of wave solutions for a scalar nonlinear diffusion equation of the form \[ \frac{\partial u}{\partial t} - \operatorname{div}( A(x) \nabla u ) = F(x,u),\quad x\in\mathbb{R}^N, \eqno{(*)} \] with arbitrary dimension \(N\). The coefficients \(A\) and \(F\) are assumed to be sufficiently smooth and \(\mathbb{Z}^N\)-periodic, i.e. \(1\)-periodic in all \(x\)-directions. Moreover, the coefficient \(A\) guarantees the uniform parabolicity of the differential operator in the left-hand side of (\(*\)). Furthermore, the nonlinearity \(F\) is assumed to be of bistable type such that the equation (\(*\)) restricted to the \(N\)-dimensional torus \(\mathbb{T}^N = \mathbb{R}/\mathbb{Z}^N\) has two stable stationary states \(\psi^- < \psi^+\) with no other stable stationary states between \(\psi^-\) and \(\psi^+\). An entire solution of equation (\(*\)) of the form \(u(x,t) = U(x\cdot e - c t, x)\), where \(U\::\:\mathbb{R}\times\mathbb{T}^N\to\mathbb{R}\), \(c\in\mathbb{R}\) and \(e\in\mathbb{S}^{N-1}\) is called the traveling wave solution connecting \(\psi^-\) and \(\psi^+\), if \[ \lim\limits_{s\to\pm\infty} U(s,x) = \psi^\pm(x) \quad\text{uniformly with respect to } x\in\mathbb{T}^N. \] Depending on the quantity \[ \mathcal{I} = \int_{\mathbb{T}^N} dx \int_{\psi^-(x)}^{\psi^+(x)} F(x,s) ds, \] the author distinguishes two cases: balanced (if \(\mathcal{I} = 0\)) and nonbalanced (if \(\mathcal{I} \not= 0\)). In the former case, he proves the existence of a standing wave solution (i.e. a travelling wave with \(c=0\)), while in the latter case he proves the existence of a wave solution, which can be either a travelling wave (\(c\not=0\)) or a standing wave (\(c=0\)). The obtained results are illustrated by their application to the two-dimensional parabolic equation \[ \partial_t u = d^2 \Delta u + r(x) u ( u - a(x) ) ( 1 - u ),\quad(x,t)\in\mathbb{R}^2\times(0,\infty), \] where the diffusion coefficient \(d\) is either very small \(d\ll 1\) or very large \(d\gg 1\), and \(a,r\in C^\gamma(\mathbb{T}^2)\), \(\gamma\in(0,1)\), such that \(0 < a(x) < 1\) and \(r(x) > 0\).