Epidemic Diffusion of Data in Opportunistic Networks

Opportunistic networks have interesting communication behavior that could enable to bring ad hoc networks to people's everyday life. To allow this to happen, several efforts are still required to provide routing strategies capable to efficiently cope with the highly variable network topology and the need to tolerate temporary network partitions. This paper focuses on this argument and proposes the use of epidemic algorithms to provide best effort data delivery to all nodes in an opportunistic scenario. The main contribution of the paper is to show that the approach actually has the ability to deal with the mentioned constraints and outperforms flooding to diffuse packets over the network. The paper presents the first set of simulation results and highlights the parameters that influence the performances of the protocol. These results represent a sort of worst case analysis of the approach and are the starting point to design an adaptive diffusion protocol which is capable to adapt to both the dynamics of the network and the mobility conditions.