The equation of symmetric Markovian random evolution in a plane (Q1805760): Difference between revisions

Let \(E_{0},\ldots ,E_{n-1}\), \(n\geq 2\), be given vectors in the plane \(\mathbb R^2\), the vector \(E_{i}\) standing at an angle \(\frac {2\pi i}{n}\) to the \(x\)-axis. Let us consider a particle moving in the plane with a constant (finite) velocity \(v\), and a homogeneous Poisson process with the intensity \(\lambda >0\). The particle moves in the direction \(E_i\) until the next jump of the Poisson process occurs, then it chooses any of the directions \(E_j\), \(j\neq i\), with probability \(\frac 1{n-1}\). Let \((X(t),Y(t))\) be the position of the particle at time \(t\), set \(F(t,x,y)=\text{Probab} \{X(t)<x,Y(t)<y\}\). Denote by \(f(t,\cdot ,\cdot)\) the density of the (necessarily nontrivial) absolutely continuous component of the transition function \(F\). It is shown that the function \(f\) satisfies a hyperbolic equation \(Hf=0\), where \[ \begin{multlined} H = \frac {\partial}{\partial t}\sum ^{[(n+1)/2]}_{k=1} (-1)^{k-1}\binom {n-k}{k-1}Z^{n-2k+1}Q^{k-1}- 2 \sum ^{[n/2]}_{k=1} (-1)^{k-1}\binom {n-k-1}{k-1}Z^{n-2k} Q^{k} \\ - 2 \left (-\frac {v}2\right)^{n} \sum ^{[n/2]}_{k=0} (-1)^{k}\binom {n}{2k} \left (\frac {\partial}{\partial x} \right)^{n-2k}\left (\frac {\partial}{\partial y}\right) ^{2k}, \quad Z\equiv \frac {\partial}{\partial t} + \frac {\lambda n}{n-1}, \;Q\equiv \frac {v^2}4\Delta , \end{multlined} \] \(\Delta \) denoting the two-dimensional Laplacian. In the case \(n=2\) this result implies the classical probabilistic representation of solutions to the telegraph equation due to \textit{S. Goldstein} [Q. J. Mech. Appl. Math. 4, 129-156 (1951; Zbl 0045.08102)] and \textit{M. Kac} [see e.g. Rocky Mt. J. Math. 4, 497-509 (1974; Zbl 0314.60052)]. Moreover, a parabolic approximation to the equation \(Hf=0\) is found for \(v\to \infty \), \(\lambda \to \infty \) satisfying \(\lambda ^{-1}v^{2}\to c>0\).