Cascading failures of wireless sensor networks

Despite the fact that the research related to the invulnerability of wireless sensor networks (WSNs) has made gratifying progress, most of it focuses on their topology tolerance but fails to consider the impact of traffic load and energy consumption on their invulnerability. In this work, we build a cascading model of WSNs by considering energy exhaustion, invalid connectivity and traffic overload. We evaluate the performance of two types of WSN models, i.e., scale-free and random network ones, when facing cascading failures. The study shows that a random network model performs better than a scale-free one. Through the evaluation of the contributions of causes to node failures, we find that although traffic load is the triggering condition for cascading failures, invalid connectivity is a major reason for node failures in the network.