Effect of Vaccination on Stability of Wireless Sensor Network Against Malware Attack: An Epidemiological Model

Wireless sensor network (WSN) has gained increased prominence due to fast deployment of sensors in IoT. With WSN being deployed in varying fields viz. industrial automation, surveillance, security in these networks has become important consideration. Security solutions employed in infrastructure networks are unsuitable for resource constrained WSNs. To limit power consumption within the node, implementing antivirus/ malware protection within the node is not recommended. However, WSNs are equally prone and vulnerable to such threats (malware, worm, virus etc.). Due to distributed topology and decentralized network control, malware detection and isolation of infected nodes needs to be carried out within the network. Techniques such as SIS (Susceptible–Infected–Susceptible), SIR (Susceptible–Infected–Recovered), SCIR (Susceptible–Carrier–Infectious–Recovered) have been proposed for malware detection in WSNs. However, the techniques suffer from quarantine and vaccination. In this paper, we propose an epidemiological model for immunizing the network against such attacks. In our method, the nodes are initially vaccinated and then deployed. This protects the malware infection after deployment. The model also has a quarantine state. A vaccinated node getting infected after deployment is placed in this state. Simulation results have shown lower probability of malware attack.