文摘
In recent years, we have witnessed an increase in the popularity of mobile wireless devices and networks, with greater attention devoted to feasibility of opportunistic computing, sensing, and communication. In Mobile Social Networks (MSNs), communication is provided by spatial proximity and social links between peers, where personal devices carried by users communicate directly in a device-to-device mode. On one hand, human mobility provides encounters between peers and opportunities for communication without additional infrastructure; on the other hand, it introduces intermittent connections, network partitions, and long delay, requiring sophisticated message-forwarding mechanisms to improve network performance. Therefore, socially-inspired approaches which consider network structure and personal user features have been proposed to cope with these challenges. However, many studies disregard adaptive policies of message forwarding capable of dealing with variations of these features. In this paper, we investigated message dissemination in MSNs considering external factors such as temperature and seasonal calendar as environmental features capable of model users’preferences and encounters. We evaluated the time of day, the day of the week, and environmental variables such as weather and geographic position as important factors to the collective behavior and spatiotemporal characteristics of urban scenarios. This paper presents an analysis of real data from weather and human mobility, which depict distinct social interactions and spatial features characterized by changes in thermal conditions. Thus, we propose a socially-aware forwarding mechanism that is adaptable to the seasonality of personal preferences. Our experiments indicated that pervasive data can provide useful information towards the design of the next generation of human-centered Opportunistic Networks.
