Vincent Danos

University of Edinburgh

Title: Statistical physics of communicating processes

Abstract:  We propose a probabilistic interpretation of a class of
reversible communicating processes. The rates of forward and backward
computing steps, instead of being given explicitly, are derived from a
set of formal energy parameters. This is similar to the
Metropolis-Hastings algorithm but in a distributed context. We find that
with the right definition of the energy landscape, a specific lower
bound on the cost of communication will guarantee that a process reaches
a probabilistic equilibrium state and hits a success state in finite
average time, if there is one.