Analysis on invulnerability of wireless sensor networks based on cellular automata

Existing failure models for wireless sensor networks (WSNs) usually focus on the impact of specific failure events on the network. However, the actual fault process of WSNs is driven by multiple fault events, and their failure behavior is more complex and unpredictable than that only considering specific events. Therefore, this paper proposes a more realistic failure model for WSNs based on cellular automata (CA), which considers energy exhaustion, hardware/software malfunctions and impaired connectivity. On this basis, we explore the network invulnerability under three attack strategies (i.e., random attacks, maximum-degree attacks and maximum-betweenness attacks). Moreover, we investigate the impact of sink node layout on network invulnerability by comparing three sink node layout schemes (i.e., random layout, degree-oriented layout and betweenness-oriented layout). Experimental results have shown that the maximum-betweenness attack strategy has the greatest damage to network invulnerability; in the face of wide-range attacks, isolation from the sink nodes is the main cause of node failures in WSNs; the betweenness-oriented layout performs better in terms of improving network invulnerability. The obtained results can provide theoretical guidance for users to build an invulnerable WSN.

Automated summary

This paper proposes a more realistic failure model for wireless sensor networks (WSNs) based on cellular automata (CA), and investigates the vulnerability of the network and the impact of sink node layout on network invulnerability. Experimental results show that the maximum-betweenness attack strategy has the greatest damage to network invulnerability, isolation from the sink nodes is the main cause of node failures in WSNs when facing wide-range attacks, and the betweenness-oriented layout performs better in improving network invulnerability.