SMS Worm Propagation Over Contact Social Networks: Modeling and Validation

Nowadays, short message service (SMS) worms have been discovered to propagate themselves via victims' contact lists by sending malicious text messages. Correspondingly, defenders need to analyze and model the dynamics of these worms to lessen their potential threat. However, the existing worm propagation models, which almost generate the similar curves of an exponential smooth rise, cannot well explain the infection dynamics of real-world SMS worms, which exhibits an uneven wave-like uplift. Motivated by this observation, we formalize the general infection process of SMS worms in contact social networks, and propose a novel analytical model based on stochastic processes. In contrast to previous models, our model not only considers the different and asymmetrical relationships between mobile users by modeling the node reputation and the edge trust degree, but also describes the user behavior of checking messages by introducing two susceptible states. Moreover, the strong assumptions in previous works are eliminated by determining related components based on extensive statistical investigations. Afterward, both real-world SMS worms and artificial ones are utilized in validation and comparison experiments, and the results show that our model is more suitable for describing the propagation of these sophisticated worms, compared with the state-of-the-art models. In addition, we study on the impacts of key factors and give some interesting discoveries.