A Network-Level Cybersecurity Risk Detection Model for Cloud-based Intelligent Transportation Systems

Abstract

Intelligent Transportation Systems (ITS) have become indispensable to modern urban environments, where they support the optimization of traffic flow, the improvement of road safety, and the efficient management of transportation infrastructure. These systems rely on a wide range of sensors, Internet of Things (IoT) devices, and networking technologies to collect large volumes of data, and they typically depend on cloud computing platforms for storage and processing. While this cloud-based architecture provides clear advantages in scalability and computational power, it also introduces significant cybersecurity concerns related to data confidentiality, integrity, and availability. This paper presents a conceptual model for ITS deployed in cloud environments, examines the principal security challenges that arise at each layer of the architecture, and discusses contemporary methods and approaches for mitigating these risks. The principal technical contribution of the paper is the Adaptive Centroid Radius Model (ASRM), a lightweight, mathematically transparent, and parameter-tunable method for real-time anomaly detection in network traffic. ASRM is validated through a numerical experiment and benchmarked against four state-of-the-art anomaly-detection methods on the NSL-KDD dataset. Experimental results demonstrate that ASRM achieves the highest precision (0.81) among all evaluated methods and reduces training and prediction times by up to three orders of magnitude relative to ensemble- and kernel-based detectors, confirming its suitability for deployment on resource-constrained edge devices.