Tracking Social Interaction via Time Varying Path Loss Estimation from Wireless Transmissions

Abstract

Social networking services, websites and mobile apps have become increasingly popular over the past years. As these services and applications become more sophisticated and encompassing, they are constantly looking for new methods of inferring social connections between peers. Some methods currently being used rely on identifying simultaneous geographical location of peers by means of the Global Positioning System and Internet Protocol addresses. These methods have limited accuracy and in most cases require actions by the peers. In this work, we consider utilizing estimations of the time-varying path loss of wireless transmissions from peers to infer their social interactions. Using a set of experiments, we demonstrate that the proposed method can be used to detect periods of social interactions by a monitoring receiver that is not coordinated with the peers.