Bootstrapping opportunistic networks using social roles

Abstract

Opportunistic routing protocols can enable message delivery in disconnected networks of mobile devices. To conserve energy in mobile environments, such routing protocols must minimise unnecessary message-forwarding. This paper presents an opportunistic routing protocol that leverages social role information. We compute node roles from a social network graph to identify nodes with similar contact relationships, and use these roles to determine routing decisions. By using pre-existing social network information, such as online social network friends, to determine roles, we show that our protocol can bootstrap a new opportunistic network without the delay incurred by encounter-history-based routing protocols such as SimbetTS. Simulations with four real-world datasets show improved performance over SimbetTS, with performance approaching Epidemic routing in some scenarios.