Design and performance of an AI-enabled threat intelligence framework for IoT-enabled autonomous vehicles

In the realm of improving living standards, the rise of Internet of Things (IoT)-enabled autonomous vehicles promises significant global economic potential, estimated to reach trillions of dollars. However, the rapid expansion of IoT-integrated vehicles, coupled with the prevalence of IoT ecosystems, introduces unprecedented challenges. Diverse threats such as information gain attacks, distributed denial of service, and persistent cyber botnet attacks pose substantial risks to IoT-integrated autonomous vehicles. Addressing this critical issue, we propose an innovative solution — an ensemble learning-based cyber threat intelligence mechanism adept at efficiently detecting sophisticated multi-variant cyber threats and attacks. To validate the effectiveness of the proposed model, we conducted extensive experiments on a publicly available state-of-the-art Kitsune dataset. We used standard performance evaluation metrics to systematically assess the mechanism's performance, including a comprehensive comparison with recently proposed ensemble and hybrid Deep Learning (DL) architectures, as well as benchmark DL-algorithms. The proposed technique demonstrates promising results, excelling in both speed efficiency and high accuracy in detecting multi-variant cyber threats. To further validate the integrity of our results, the proposed mechanism was subjected to rigorous cross-validation, establishing its unbiased performance and reliability. This research not only contributes to the evolving landscape of IoT security but also provides a robust and efficient solution to the escalating challenges faced by autonomous vehicles in the interconnected digital age.