Delay-tolerant networking (DTN) is a network architecture characterized by the lack of continuous connectivity. Messages are delivered by moving nodes in a store-and-forward manner. In such a network, the mobility models of nodes play an important role in DTNs, because messages can only be delivered when two or more nodes contact each other. In general, mobility models can be categorized into synthetic and trace models. Synthetic models are based on mathematical models that generate the mobility models. Trace models record people's daily movements in the real world; these models faithfully render the actual situation of people's movements in their lives. However, there is a challenge in studying human mobility, and the mobility models of humans affect the performance of routing protocols in DTNs. In this article, we design an Android application to collect the mobility traces of college students in a campus environment, called NCCU Trace Data. We design a mobility model that traces students' movement, and this model can be imported into the ONE simulator to verify routing protocols. More importantly, it can be used to evaluate the performance of a social-based routing method. Finally, we evaluate our routing method and compare it to other routing methods in each mobility model. The simulation results show that our newly designed mobility model is a step closer to the real environment.