Quantitative survivability analysis using probability model checking: A study of cluster-based vehicle networks with dual cluster heads

Abstract

As an important part of future 5G wireless networks, a vehicular network demands safety, reliability and connectivity. In this context, networking survivability is usually considered an important metric to evaluate network performance. In this paper, we propose a survivability model for vehicle communication networking based on dual cluster heads, wherein a backup cluster head (CH) will be activated if the primary CH fails, thereby effectively enhancing the network lifetime. Additionally, we introduce a software rejuvenation strategy for the prime CH to further improve the survivability of the entire network. Using the Probabilistic Symbolic Model Checker (PRISM), we verify and discuss the proposed survivability model via numerical simulations. The results show that network survivability can be effectively improved by introducing an additional CH and further enhanced by adopting the software rejuvenation technique.