Characterizing User Mobility in Second Life

Abstract: In this work we present a measurement study of user mobility in Second Life. We first discuss different techniques to collect user traces and then focus on results obtained using a crawler that we built. Tempted by the question whether our methodology could provide similar results to those obtained in real-world experiments, we study the statistical distribution of user contacts and show that from a qualitative point of view user mobility in Second Life presents similar traits to those of real humans. We further push our analysis to line of sight networks that emerge from user interaction and show that they are highly clustered. Lastly, we focus on the spatial properties of user movements and observe that users in Second Life revolve around several point of interests traveling in general short distances. Besides our findings, the traces collected in this work can be very useful for trace-driven simulations of communication schemes in delay tolerant networks and their performance evaluation.