Enhancing IoT Security With Trust Management Using Ensemble XGBoost and AdaBoost Techniques

As next-generation networking environments become increasingly complex and integral to the fabric of digital transformation as the traditional perimeter-based security model proves inadequate. The Zero Trust framework emerges as a critical solution to this challenge, advocating for a security model that assumes no implicit trust and requires verification at every step. The rapid growth of the Internet of Things (IoT) creates an environment of millions of interacting devices that radically transform today's digital environment. In such conditions, the problem of identifying malicious and compromised nodes among IoT devices becomes mandatory to maintain trustworthy environment. The main objective of this research is to implement an advanced trust management mechanism, with the main transformation on the framework of security within which IoT environments become workable. To overcome this, an approach based on intelligent ensemble learning is presented in this paper by formulating a dataset consists of IoT-23, N-BaIoT, Edge-IIoTset, and AutoTrust-IoTDS. This work is specialized by selecting a diversified combination of features from these datasets to construct an effective dataset that would further enhance inculcation of trust in the IoT networks. The study proposed two notable models, i.e., XGBoost with logistic regression and AdaBoost with decision tree to achieve the efficiency and scalability of the proposed models. The results shows significant improvement in the efficiency, and having both the mentioned models to offer a precision, recall and F1-score of 0.99 representing efficient mechanism in ensuring a secure and dependable IoT ecosystem.