Interest-based Epidemic Routing in Opportunistic Mobile Networks

Message delivery in opportunistic mobile networks is a challenging task since the network topology constantly changes and end-to-end paths can hardly be sustained. Epidemic routing forwards a copy message to each contacted node to achieve a high network delivery performance; this however easily burdens the network nodes with high traffic load, quickly depleting the node's resources, e.g. power and storage, and finally degrading the network delivery performance. This paper proposes an interest-based Epidemic that improves Epidemic to be a content-aware forwarding by taking message content, node interest, and node community into consideration. Using simulation, driven by real human contact datasets, we investigate the performance of the proposed algorithm compared with Epidemic (content-oblivious) and Direct Transmission (content-aware), in terms of total delivered messages, average convergence time, and total relayed messages. Simulation results show that Epidemic-Interest outperforms Direct Transmission in terms of total delivered message and average convergence time. Moreover, compared with Epidemic, it can reduces the transmission cost while keeping the total delivered messages as high as Epidemic's; however, it increases the convergence time beyond that of Epidemic.