A Location Prediction Scheme Based on Social Correlation

With the development of mobile networks, terminals´ locations become increasingly valuable for both the whole system performance and terminals´ applications or services. As knowing the knowledge of terminals´ location information in advance would provide extern preparation time to accelerate response to terminals´ requirements, location prediction plays a quite important role. Since most of the mobile terminals are human-carried devices, terminals´ location prediction is actually the persons´ location prediction. On the other hand, the research on social networks has experienced a fast development and provided a new kind of view on prediction. In our paper, we propose a location prediction scheme based on the idea of social correlation which is a kind of link between different terminals weighted by statistic results of history information.We also model topologies as sequences of snapshots to guide prediction. Few limits on the format of history location information record makes sure a strong adaptiveness of our scheme to various application´s requirements.