Invulnerability Analysis of Wireless Sensor Networks based on Cellular Automata

Network invulnerability is widely considered as one of the main bottlenecks restricting the widespread application of wireless sensor networks (WSNs). Therefore, in this work, we present a 2-D cellular automata model to mimic the failure process of WSNs. In this model, we consider five failure events of sensor nodes (i.e., energy exhaustion, hardware/software malfunctions, intentional attacks, random attacks and isolation from the sink node). The experimental results have shown that the invulnerability of WSNs under random attacks is better than that under intentional attacks. In addition, through composition analysis of failed sensor nodes, we observe that energy exhaustion is mainly responsible for node failures in the case of random attacks. Isolation from the sink node is the main cause of node failures under intentional attacks. The obtained results can provide theoretical guidance to build a more invulnerable WSN.