Analysis and design of mobile Wireless Social Model

Delay Tolerant Networks (DTNs) are characterized by very long delay paths and frequent network partitions. Most existing protocols are based on simplistic models which differ from real scenarios or real traces which do not allow for sensitivity analysis. In this paper, we analyze the differences between the Social Network (SN) and the Wireless Social Network (W-SN) in terms of the clustering coefficient, characteristic path length, community size, and membership number. We observe that the clustering coefficient, community size and membership number in the W-SN are bigger than those in the SN, while the character path length in W-SN becomes smaller. And then we propose a new Wireless Social Model (WSM) based on community character. Node in WSM is driven by the α-Model and has an equal probability to move to one of his adjacent vertices. We evaluate our model via real traces supplied by Dartmouth College, which shows that our wireless social mobile model offers a good approximation of real movements.