Trust-Aware AI-Enabled Edge Framework for Intelligent Traffic Control in Cyber-Physical Systems

Abstract

The rapid evolution of smart cities has led to the deployment of Cyber-Physical IoT Systems (CPS-IoT) for real-time monitoring, intelligent decision-making, and efficient resource management, particularly in intelligent transportation and vehicular networks. Edge intelligence plays a crucial role in these systems by enabling low-latency processing and localized optimization for dynamic, data-intensive, and vehicular environments. However, challenges such as high computational overhead, uneven load distribution, and inefficient utilization of communication resources significantly hinder scalability and responsiveness. Our research presents a robust framework that integrates artificial intelligence and edge-level traffic prediction for CPS-IoT systems. Distributed computing for selecting forwarders and analyzing threats across the IoT system enhances stability while improving energy efficiency. In addition, to achieve efficient routing decision-making, the Artificial Bee Colony algorithm is explored to enhance the effective utilization of network resources across IoT systems. Based on the simulation results, the proposed framework achieves remarkable performance in terms of throughput by 38%–41%, packet loss ratio by 30%–33%, security risk mitigation by 35%–37%, and trust level by 41%–44% as compared to existing work.