Cleaning an Arbitrary Regular Network with Mobile Agents

In this paper, we consider a contaminated network with an intruder. The task for the mobile agents is to decontaminate all hosts while preventing a recontamination and to do so as efficiently as possible. We study under what conditions and what cost a team of mobile agents can do this in synchronous arbitrary regular graphs using the breadth-first-search strategy. Due to the nature of the experiment we use a genetic algorithm to find the minimum number of agents required to decontaminate a given network. The results show that there is a relation between the degree, the size of the graph, and the number of starting locations of the mobile agents. in particular, this relation demonstrates the possibility of improvements in reducing the number of mobile agents used depending on the number of starting location in arbitrary regular graphs.