Exploring seasonal human behavior in opportunistic mobile networks

In recent years, there is a growing research interest in smart city applications based on opportunistic communications. The opportunistic networks in these scenarios are composed of disconnections, network partitions, high delay and strong influence of human mobility. To cope with these challenges, social-inspired approaches have been proposed considering the structure of networks and personal user features, however, limited studies have explored temporal variations of features and influence of exogenous variables, disregarding adaptive forwarding policies recommended for dynamic scenarios. In this paper, we address these challenges while investigating both, the temperature and the season calendar, as environmental features able to model the behavior of users' mobility and peer contacts. The results showed distinct social and spatiotemporal features characterized by thermal conditions able to affect the network performance. We also identified critical points of temperature able to provide early signals about the network changes. Finally, our results indicate that environmental data are crucial information towards the design of the next generation opportunistic mobile networks.