Performance evaluation of WiFi Direct for data dissemination in mobile social networks

WiFi Direct is a recent device-to-device communication technology standardized by the WiFi Alliance. Its increasing availability on popular mobile systems (e.g. Android) presents a unique opportunity for developers to implement mobile social networks (MSNs), a new paradigm that facilitates data dissemination without Internet access by leveraging human mobility and short-range communication technologies. Since WiFi Direct is not originally designed for such applications, it is significant to learn its performance in practice. In this paper, we investigate goodput and fairness of WiFi Direct for data dissemination in MSNs. To this end, we develop an MSN application and conduct three sets of experiments on a testbed comprising several Android devices. Experimental results show that the data loads and mobility of nodes greatly impact the goodput and fairness.