Research of Optimal Task Scheduling for Distributed Real-time Embedded Systems

Abstract

Distributed real-time embedded (DRE) systems are becoming increasingly common and important as the underlying technology for distributed computing and networking systems continues to develop. Efficient task scheduling is essential for achieving high performance in the application of embedded systems. However, because of its NP-hard nature, there are mainly heuristic algorithms in this area still needed to be improved. In this paper, a modified dynamic critical path algorithm is presented to find the earliest possible start time and the latest possible finish time of a task using the distributed nodes network structure. An algorithm called the combined dynamic BLevel (CBL) has been presented to select the node-task mapping priorities based on the dynamic critical path. Using a novel DRE network model and a communication scheduling algorithm, tasks and associated information events are scheduled simultaneously with the transfer time onto a nodepsilas communication link. According to the simulation results, the proposed solution outperforms existing mechanisms in terms of schedule length for communication intensive tasks.