A Machine Learning Framework for Cyberattack Detection and Risk Impact Assessment in IoT Use Cases

Internet of things (IoT) technology has revolutionized industrial applications leading to Industrial IoT (IoT). In such distributed environments, there is high probability of cyberattacks due to heterogeneity of devices, protocols and techniques. Artificial intelligence (AI), of late, became instrumental in enhancing security of different applications. However, existing methods focused on attack or intrusion detection but there is little exploration on risk impact analysis. In this paper, we propose a machine learning (ML) framework which has a pipeline of operations achieving multiple objectives such as intrusion detection, attack classification and also risk impact assessment. Our framework, named IntelliSec is designed to have more comprehensive approach in securing digital infrastructure in IoT environments. We proposed two algorithms namely Intelligent Intrusion Detection and Classification (IIDC) and Risk Impact Analysis (RIA) to realize IntelliSec functionality. The former exploits learning based phenomena to acquire intelligence incrementally and detect intrusions from time to time and classify them while the latter analyses risk impact on the system. Thus the proposed framework takes care of intelligent means of protecting cyberspace and enable security professionals with a reliable tool. Our empirical study has revealed that IntelliSec has its potential to leverage security of IoT.