Decontamination with Temporal Immunity by Mobile Cellular Automata

In this paper we consider the network decontamination problem in a mobile cellular automata (MCA). The system consists of a two-dimensional lattice that evolves like a cellular automata, where however some cells are in a special active state. Such a state indicates the presence of an agent which also follows a local transition rule to move from cell to neighbouring cell. A dynamic contamination process causes the spread of a virus (or a fault), and the presence of an agent on a cell guarantees local disinfection (or decontamination). Once disinfected, a cell stays immune to recontamination for a predetermined amount of time. The goal is to design the local rules for the agents and their initial placement so that the agents can decontaminate the entire system without allowing any cell to be re-contaminated. To be efficient, the decontamination should employ as few agents as possible. We design several strategies depending on the type of neighborhood, and on the ability of the agents to clone themselves.