Human mobility model based on time-varying bipartite graph

Nowadays human beings are surrounded by a heterogeneous networking environment consisting a growing number of portable computation and communication devices. As most devices are carried out by human beings, such a contact-based networks is highly influenced by human mobility. This fact implies that the presence of possible patterns in human movements can be exploited by wireless network applications in order to extract sensible informations on top of which novel mobile services can be deployed. Such information does not cover only the spatial or temporal dimension, but also concerns relational and social aspects of the involved people. In order to evaluate such applications we have to develop a mobility simulation sufficiently expressive and easily tunable. The most important goal in the mobility model research area is to provide a tool that can capture the most important and relevant features regarding both physical and social dimensions. For my PhD research I propose a new mobility model able to properly reproduce the spatial, temporal and social features that can be observed in real mobility datasets. In the model people move within a set of geo-communities, i.e. locations loosely shared among people, according to a bipartite time-varying graph; similarly, inside a geo-community, people move according to a modified version of classical random waypoint. We also derive social relationships from the bipartite graph representation by means of different types of projections on the node set. The purpose of this document is to briefly describe the state of the art in mobility model, and to outline my planned PhD research.