Skip to Main Content
The complexity of social mobile networks, networks of devices carried by humans (e.g. sensors or PDAs) and communicating with short-range wireless technology, makes it hard protocol evaluation. A simple and efficient mobility model such as SWIM reflects correctly kernel properties of human movement and, at the same time, allows to evaluate accurately protocols in this context. In this paper we investigate the properties of SWIM, in particular how SWIM is able to generate social behavior among the nodes and how SWIM is able to model networks with a power-law exponential decay dichotomy of inter contact time and with complex sub-structures (communities) as the ones observed in the real data traces. We simulate three real scenarios and compare the synthetic data with real world data in terms of inter-contact, contact duration, number of contacts, and presence and structure of communities among nodes and find out a very good matching. By comparing the performance of BUBBLE, a community-based forwarding protocol for social mobile networks, on both real and synthetic data traces, we show that SWIM not only is able to extrapolate key properties of human mobility but also is very accurate in predicting performance of protocols based on social human sub-structures.