AI-Driven Cybersecurity Situational Awareness for IoT Networks: Enhancing Threat Detection and Prediction with Machine Learning

With the rapid development of network technologies, emerging innovations such as the Internet of Things (IoT), cloud computing, and big data are becoming increasingly widespread. However, this proliferation has also intensified cybersecurity threats in cyberspace. As a result, cyber security situational awareness (CSA) has attracted significant attention from both academia and industry. Typically, CSA models comprise three key levels: element extraction, situational understanding, and situational prediction. Among these, situational prediction which involves forecasting the overall security posture of networks holds critical theoretical and practical value. This paper investigates how artificial neural networks and inverse transactions can be effectively leveraged to enhance cybersecurity situational awareness. By integrating these advanced technologies, the study aims to improve predictive and responsive capabilities against cyber threats, thereby strengthening network defense systems and addressing the ever-evolving landscape of cyber risks. To address the gap in theoretical modeling for CSA, a novel cybersecurity situational awareness framework is proposed. Its effectiveness is evaluated across several dimensions, including model structure, knowledge representation, and assessment methods. Improved two-party and three-party authentication schemes tailored for IoT environments are introduced, alongside a machine learning-based authentication framework. These approaches, combined with enhanced identification techniques, outperform traditional methods in terms of accuracy. Furthermore, the study explores the application of these methods in key areas such as security, data transmission, system survivability, and comprehensive system assessment showcasing the latest advancements and outlining future directions in CSA research.