Towards Automation of Vulnerability and Exploitation Identification in IIoT Networks

Since Industrial Internet of Things (IIoT) networks are comprised of heterogeneous manufacturing and technological devices and services, discovering previously unknown vulnerabilities and their exploitation vectors (also known as Penetration Testing - PT) is an arduous and risk-prone process. PT across IIoT networks requires system administrators to attempt multiple and often bespoke commercial tools for testing vulnerable network nodes, platforms, and software. In this paper, we propose a new testbed IIoT environment involving multiple vulnerable platforms connected to IIoT sensors and IoT gateways for designing automated vulnerability and exploitation identification techniques based on analyzing network flows. We utilize a particle filter technique for estimating the vulnerability and exploitation behaviors in a term of posterior probabilities. The proposed model is better than using traditional artificial planning algorithms that consume significant computational resources and demand termination criteria. The proposed testbed IIoT environment can be shared with other like-minded researchers to facilitate future evaluations.