Edge-Cloud Latency Optimization: A Priority-Based Scheduling Framework with Dynamic VM Allocation

The proliferation of real-time data from Internet of Things (IoT) devices has posed significant challenges to traditional cloud-centric resource management. In particular, latency constraints and overload issues arise when large volumes of sensor data are continuously transmitted to remote cloud servers for processing. In this paper, we propose a novel framework that integrates priority-based data scheduling with dynamic virtual machine (VM) allocation in edge-cloud environments. Our approach employs an integrated flow queue at the edge node to eliminate inefficiencies associated with multi-queue systems. Three scheduling algorithms (static priority, dynamic priority, and integrated priority) are presented to classify and process data based on urgency and sensor type. The edge node handles high-priority tasks locally, while time-tolerant tasks are delegated to cloud VMs. Using real-world sensor data—including heartbeat and body temperature—we demonstrate that our proposed model significantly reduces latency and optimizes resource utilization compared to existing solutions. Our results suggest that this framework can be a scalable and robust solution for emerging IoT applications requiring real-time data processing.