An enhanced community-based mobility model for distributed mobile social networks

摘要

Simulation is fundamental tool for the evaluation and validation of the applications and protocols in Mobile Social Networks. However, the limited number of real user traces available and the imposed restrictions of the specific scenarios, make generalization very hard. Therefore, the need has been created for synthetic mobility models. The widely used Random Way-Point Mobility Model has been proven unable to capture characteristics of human mobility such as the social attraction. Consequently, in recent years mobility models based on social network theory, able to capture the temporal and spatial dependencies of mobile social networks, are being designed. In this paper the Enhanced Community Mobility Model (ECMM) is introduced. It follows preceding community-based approaches, that map communities to a topological space. Its main contribution is the introduction of new features, such as pause periods and group mobility encouragement, lacking for previous community-based mobility models. Additionally, ECMM enables researchers to arbitrarily select a social model as the trace generation process input, while at the same time generates traces with high conformance to that social network. A comparison between synthetic traces, generated by ECMM, other community-based models and a number of real ones is provided for validation.