Performance analysis of epidemic routing in Dtns with limited forwarding times and selfish nodes

摘要

Routing protocols in Delay-Tolerant Networks (Dtns) often need nodes serving as relays to carry and forward the messages. Because of the energy constraint, each node cannot forward too many times. Moreover, the maximal forwarding limes may be heterogeneous for different nodes. In addition, nodes can form many communities according to certain social relations, and nodes in the same community can be seen as friends. Because of the selfish nature, one node may use its limited forwarding times to help its friends. This paper proposes a theoretical framework to evaluate the performance of the Epidemic Routing (ER) algorithm in Dtns in such complex environment for the first time. Simulations based on both synthetic and real motion traces show the accuracy of our framework. Numerical results show that ER algorithm is very robust to the forwarding times. Moreover, the impact of the selfish nature is related with many factors (e.g., number of nodes, number of communities, etc.).