Broadcasting in dynamic Sierpinski and butterfly networks

Abstract

This thesis studies the broadcasting problem in dynamic networks, i.e., a network in which some edges may be disconnected and then reappear over time, still main- taining the network connected. The problem consists of broadcasting the message from a unique source, an agent, to a set of remaining mobile agents located in distinct nodes of the network. There is also an adversary that wants to prevent broadcasting by modifying the availability of the edges in order to change the configuration of the network, but maintaining the graph always connected. We considered as topology the Sierpinski graph. We firstly prove the lower bounds to this problem i.e., we study the minimum number of agents necessary for the source of the message to broadcast it to all the other agents. Then, we prove an upper bound, i.e., we propose an algorithm that solves the problem, providing a number of ignorant agents that are sufficient to solve it. Then, we analyze the lower bound to broadcasting on an n-dimensional butterfly network and we com- pare the results to the those obtained for the Sierpinski graphs.