Opportunistic Data Dissemination Using Real-World User Mobility Traces

Opportunistic communication allows humans equipped with mobile devices to exchange information via a wireless link whenever they are nearby. This work examines the performance of a profile-based data dissemination scheme. We use a 2- step simulation approach which combines realistic user mobility traces with synthetic mobility models. We consider different system configurations and evaluate their effects on the data dissemination scheme. In particular, we look at how different wireless networking technologies affect the system and how much can we benefit from the installation of fixed information sprinkler nodes. Furthermore, we investigate the effects of user behavior by comparing selfish and generous users. Our results clearly demonstrate that realistic user mobility patterns result in a feasible opportunistic data dissemination scheme. In most of the studies scenarios, the effectiveness of the dissemination is very high, with over 90% of the users reached within a week of real time. Our results on user behavior show that although the effects of selfish behavior can be mitigated, they cannot be completely eliminated.