Stationary blocking measures for one-dimensional nonzero mean exclusion processes. (Q1872288)

The exclusion processes constitute one of the main families of stochastic processes in the area of interacting particle systems. The exclusion process \(\eta_\bullet = (\eta_t)_{t\geq0}\), with random walk kernel \(p(\cdot)\), is a continuous time Markov process on \(\{0,1\}^{Z^d}\), formally defined as the Feller process with generator \[ \Omega f(\eta ) = \sum_{x,y\in Z^d} (f(\eta_{xy})-f(\eta ))p(y-x)\eta (x) (1-\eta(y)) \] for any cylinder function \(f\), where \(\eta_{xy}(x) = \eta (y)\), \(\eta_{xy}(y) = \eta (x)\), and \(\eta_{xy}(z) = \eta (z)\), for \(z\neq x,y\). A basic problem is the characterization of stationary measures for the exclusion processes, where the Bernoulli measures \(\rho_\alpha\) with densities \(\alpha, \alpha\in [0,1]\), are known to be the extremal translation invariant stationary measures for the processes with irreducible random walk kernel \(p(\cdot)\). In \(d=1\), non-translation invariant stationary measures are known to exist for specific \(p(\cdot)\). These measures are concentrated on configurations that are completely occupied far enough to the right and completely empty far enough to the left; that is, they are blocking measures. In the present paper the stationary blocking measures are proved to exist for all exclusion processes in \(d=1\), with \(p(\cdot)\) having finite range and \(\sum_x x p(x)>0\).