Leveraging Digital Twins for Anomaly Detection and Adaptive Healing in Software-defined IoT

Unusual traffic patterns produced by different IoT sensors in smart city systems can impede data flow and interfere with intelligent services. Several novel approaches have been developed using artificial intelligence techniques to identify these anomalies in the underlying network. Nevertheless, there aren’t many ways to stabilize or mitigate the network once an anomaly has been found. This paper proposes an intelligent anomaly detection technique for a digital twin-enabled Software-defined Internet of Things (SDN-IoT) environment. By using the capabilities of software-defined networks, the network can be reconfigured according to the rules that were specified by the controller. These rules are first implemented in the digital twin environment to evaluate their effects on the physical system. A small-scale testbed is deployed to gather data, which is subsequently utilized to train and evaluate the anomaly detection model in order to validate the proposed approach. The results showed the effectiveness and performance of the suggested scheme once it was validated on the created testbed.