Modelling and analysis of communication traffic heterogeneity in opportunistic networks

In opportunistic networks, direct communication betweenmobile devices is used to extend the set of services accessible through cellular or WiFi networks. Due to their key role, mobility patterns and their impact in such networks have been extensively studied. In contrast, homogeneous communication traffic between nodes is assumed in most studies. This assumption

is not generally true, as the mobility and social characteristics of nodes might affect the traffic exchanged between them. In this paper, we consider heterogeneous traffic patterns, propose appropriate models, and analyse the joint effect of traffic and mobility heterogeneity on the performance of popular forwarding algorithms. For example, we show that an increasing amount

of (traffic and/or mobility) heterogeneity renders simple schemes, like direct transmission, significantly more useful than normally considered and diminishes the added value of additional randomly sprayed copies. We further validate these findings on datasets collected from real networks.