Dynamic Dependent Task Scheduling for Real-Time Multi-edge-node Collaboration Computing

The rapid development of the Internet of Things generates massive data volume, placing unprecedented demands on real-time task processing. This remains a significant challenge in effectively scheduling and offloading dependent tasks while considering limited computational resources and network bandwidth constraints to minimize task completion time. In this paper, we propose a novel Dynamic Edge Heterogeneous Earliest-Finish-Time (DE-HEFT) algorithm to address the dependent task scheduling and computation offloading challenges in a multi-edge-node collaborative environment. The DE-HEFT’s time and space complexities are implemented by O(N logN) and O(N<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup>), respectively. The DE-HEFT supports real-time task migration and re-scheduling, allowing it to dynamically adapt to changes of edge node resources. Experimental results demonstrate that, compared to leading baselines, our DE-HEFT significantly reduces system latency and achieves an average delay reduction of 12.65%. This study provides a cutting-edge, efficient, and reliable solution for effective dependent task scheduling in multi-edge-node collaborative computing environments.