An IoT-based bag-of-tasks scheduling framework for deadline-sensitive applications in fog-cloud environment

The Internet of Things (IoT) technology has become a transformative force in both information and industrial sectors, enabling devices to collect and exchange data. The exponential growth of IoT devices has led to increased big data generation on the one hand and processing demands on the other. Efficient and scalable solutions are needed to handle the complexities associated with IoT systems’ deployments. In this context, scheduling IoT applications that have a special sensitivity to latency pose greater challenges. This research paper focuses on scheduling IoT Bag of Task (BoT) applications in a hybrid fog-cloud environment, with a particular emphasis on hard deadline constraints in latency-sensitive systems (such as IoT healthcare and monitoring systems). The proposed scheduling algorithm dynamically allocates BoTs on the fog platform, forwarding them to the cloud layer if the fog layer cannot meet the application’s deadline. The algorithm considers multiple types of cloud instances to meet processing requirements and minimize execution costs, while ensuring compliance with deadlines. The experimental results reveal that the proposed scheduling algorithm prevents any deadline violations. Additionally, the proposed algorithm maximizes the utilization of processing resources within fog-cloud environments, resulting in minimized execution costs.