Self-management of Routing on Human Proximity Networks

Many modern network applications, including sensor networks and MANETs, have dynamic topologies that reflect processes occurring in the outside world. These dynamic processes are a challenge to traditional information dissemination techniques, as the appropriate strategy changes according to the changes in topology. We show how network dynamics can be exploited to design a self-organising data dissemination mechanism using only node-level (local) information, which detects and adapts to periodic patterns in the network topology. We demonstrate our approach against real-world human-proximity networks.