| Resumo: | Nowadays human mobility modeling is gaining increasing attention from the research community because of its importance in wireless networks, particularly for message delivery in infrastructure-less networks such as Delay Tolerant Networks and Opportunistic Networks. Understanding the role and potential of mobile devices as relaying nodes in message dissemination and delivery depends on the knowledge about encounter patterns among nodes. Data about the usage of Wi-Fi networks can be used to perform an analysis of encounters between pairs of mobile devices, and then be extrapolated for other contexts. One of the main issues in extracting realistic encounters through Wi-Fi traces is the occurrence of ping-pong events that, if ignored, might lead to unrealistic encounter patterns and mislead conclusions. So it is important to deal with this challenge for extracting realistic encounter patterns among mobile nodes. In this paper we evaluate the impact of ping-pong events on main connectivity properties of node encounters such as Contact Time (CT) and Inter Contact Time (ICT) at both individual pairs and aggregate levels. We observed that the impact of ping-pong events at the aggregate level is not pronounced while having significant impact on ICT of some pairs of nodes.
|
|---|