Work-in-Progress: Federated and Bundled-Based DAG Scheduling

In the rapidly evolving landscape of real-time systems, particularly autonomous-driving technologies, directed acyclic graphs (DAGs) have emerged as an essential model for delineating task dependencies and facilitating parallel processing. This growing complexity and computational demand are making it increasingly challenging to efficiently schedule tasks on multi-core processors. Among various DAG scheduling algorithms, federated scheduling stands out for its efficacy in managing tasks with implicit deadlines. However, it comes at the cost of potentially over-consuming core resources and diminishing the overall schedulability of task set. To address these challenges, this paper proposes a hybrid scheduling method tailored for sporadic DAG tasks, which integrates principles from bundled scheduling to optimize core utilization and enhance system schedulability.