Invulnerability of Clustering Wireless Sensor Networks Against Cascading Failures

Abstract

Despite the fact that current research related to the invulnerability of wireless sensor networks (WSNs) against cascading failures has made some progress, most of it focuses on the peer-to-peer structure, which is far from the reality. In this paper, we build a cascading model of clustering WSNs by introducing the concept of sensing load and relay load. We discuss the impacts of model parameters on network invulnerability and evaluate the invulnerability performance of two types of WSN topologies, i.e., scale-free network and random network. Simulation and theoretical results show that the network invulnerability is negatively related to the proportion of cluster heads and positively related to the allocation coefficient. When the degree of each sensor node bears a linear relationship with its initial load, the network invulnerability is strongest. Moreover, to alleviate the damage level caused by cascading failures, we design capacity-expansion schemes and indicate the recommended model parameter settings for each scheme.