Abstract
The rapid increase in vehicle numbers, particularly in densely populated areas, leads to significant issues such as traffic congestion, time and fuel wastage during the search for parking spots, and elevated emissions of toxic gases from idling vehicles, contributing to increased pollution. Smart parking systems, leveraging advanced Internet of Things (IoT) technologies commonly seen in Industrial Internet of Things (IIoT) applications, offer a solution to these problems by reducing traffic congestion, lowering environmental pollution, and conserving fuel. Although cloud-based parking systems have been proposed, they often suffer from high latency and bandwidth requirements due to high network usage. Fog Computing presents a more effective alternative. This paper proposes a fog computing-based smart parking system designed to address these issues for SDU University. By providing real-time information about available parking spots with minimal latency, the system optimises space usage and guides drivers efficiently, particularly during peak hours, thereby reducing the need for parking area expansion. A network usage reducing approach was developed and tested using iFogSim2 toolkit, with network usage metric for a comparative analysis between fog and cloud environments. A real case study from SDU University Park was employed to validate the system. The results demonstrate that the fog-based system outperforms the cloud-based approach in terms of efficiency and responsiveness.